Update app.py

app.py

@@ -1,154 +1,167 @@
+import os
 import gradio as gr
-import numpy as np
-import random
-
-# import spaces #[uncomment to use ZeroGPU]
-from diffusers import DiffusionPipeline
 import torch
+from diffusers import Lumina2Pipeline
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+# Set up environment
+os.environ['CUDA_VISIBLE_DEVICES'] = "0"
+
+# Load models
+def load_models():
+    model_name = "/mnt/petrelfs/zhaoshitian/models/LeX-Enhancer-full"
+    
+    model = AutoModelForCausalLM.from_pretrained(
+        model_name,
+        torch_dtype="auto",
+        device_map="auto"
+    )
+    tokenizer = AutoTokenizer.from_pretrained(model_name)
+    
+    pipe = Lumina2Pipeline.from_pretrained(
+        "/mnt/hwfile/alpha_vl/qilongwu/checkpoints/LeX-Lumina", 
+        torch_dtype=torch.bfloat16
+    )
+    pipe.to("cuda")
+    
+    return model, tokenizer, pipe
 
-device = "cuda" if torch.cuda.is_available() else "cpu"
-model_repo_id = "stabilityai/sdxl-turbo"  # Replace to the model you would like to use
-
-if torch.cuda.is_available():
-    torch_dtype = torch.float16
-else:
-    torch_dtype = torch.float32
-
-pipe = DiffusionPipeline.from_pretrained(model_repo_id, torch_dtype=torch_dtype)
-pipe = pipe.to(device)
-
-MAX_SEED = np.iinfo(np.int32).max
-MAX_IMAGE_SIZE = 1024
-
-
-# @spaces.GPU #[uncomment to use ZeroGPU]
-def infer(
-    prompt,
-    negative_prompt,
-    seed,
-    randomize_seed,
-    width,
-    height,
-    guidance_scale,
-    num_inference_steps,
-    progress=gr.Progress(track_tqdm=True),
-):
-    if randomize_seed:
-        seed = random.randint(0, MAX_SEED)
+model, tokenizer, pipe = load_models()
 
-    generator = torch.Generator().manual_seed(seed)
+def generate_enhanced_caption(image_caption, text_caption):
+    """Generate enhanced caption using the LeX-Enhancer model"""
+    combined_caption = f"{image_caption}, with the text on it: {text_caption}."
+    instruction = """
+Below is the simple caption of an image with text. Please deduce the detailed description of the image based on this simple caption. Note: 1. The description should only include visual elements and should not contain any extended meanings. 2. The visual elements should be as rich as possible, such as the main objects in the image, their respective attributes, the spatial relationships between the objects, lighting and shadows, color style, any text in the image and its style, etc. 3. The output description should be a single paragraph and should not be structured. 4. The description should avoid certain situations, such as pure white or black backgrounds, blurry text, excessive rendering of text, or harsh visual styles. 5. The detailed caption should be human readable and fluent. 6. Avoid using vague expressions such as "may be" or "might be"; the generated caption must be in a definitive, narrative tone. 7. Do not use negative sentence structures, such as "there is nothing in the image," etc. The entire caption should directly describe the content of the image. 8. The entire output should be limited to 200 words.
+"""
+    messages = [
+        {"role": "system", "content": "You are Qwen, created by Alibaba Cloud. You are a helpful assistant."},
+        {"role": "user", "content": instruction + "\nSimple Caption:\n" + combined_caption}
+    ]
+    text = tokenizer.apply_chat_template(
+        messages,
+        tokenize=False,
+        add_generation_prompt=True
+    )
+    model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
 
+    generated_ids = model.generate(
+        **model_inputs,
+        max_new_tokens=1024
+    )
+    generated_ids = [
+        output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
+    ]
+
+    response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
+    enhanced_caption = response.split("</think>", -1)[-1].strip(" ").strip("\n")
+    
+    return combined_caption, enhanced_caption
+
+def generate_image(enhanced_caption, seed, num_inference_steps, guidance_scale):
+    """Generate image using Lumina2Pipeline"""
+    generator = torch.Generator("cpu").manual_seed(seed) if seed != 0 else None
+    
     image = pipe(
-        prompt=prompt,
-        negative_prompt=negative_prompt,
+        enhanced_caption,
+        height=1024,
+        width=1024,
         guidance_scale=guidance_scale,
         num_inference_steps=num_inference_steps,
-        width=width,
-        height=height,
+        cfg_trunc_ratio=1,
+        cfg_normalization=True,
+        max_sequence_length=256,
         generator=generator,
+        system_prompt="You are an assistant designed to generate superior images with the superior degree of image-text alignment based on textual prompts or user prompts.",
     ).images[0]
-
-    return image, seed
-
-
-examples = [
-    "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k",
-    "An astronaut riding a green horse",
-    "A delicious ceviche cheesecake slice",
-]
-
-css = """
-#col-container {
-    margin: 0 auto;
-    max-width: 640px;
-}
-"""
-
-with gr.Blocks(css=css) as demo:
-    with gr.Column(elem_id="col-container"):
-        gr.Markdown(" # Text-to-Image Gradio Template")
-
-        with gr.Row():
-            prompt = gr.Text(
-                label="Prompt",
-                show_label=False,
-                max_lines=1,
-                placeholder="Enter your prompt",
-                container=False,
-            )
-
-            run_button = gr.Button("Run", scale=0, variant="primary")
-
-        result = gr.Image(label="Result", show_label=False)
-
-        with gr.Accordion("Advanced Settings", open=False):
-            negative_prompt = gr.Text(
-                label="Negative prompt",
-                max_lines=1,
-                placeholder="Enter a negative prompt",
-                visible=False,
+    
+    return image
+
+def run_pipeline(image_caption, text_caption, seed, num_inference_steps, guidance_scale):
+    """Run the complete pipeline from captions to final image"""
+    combined_caption, enhanced_caption = generate_enhanced_caption(image_caption, text_caption)
+    image = generate_image(enhanced_caption, seed, num_inference_steps, guidance_scale)
+    
+    return {
+        "combined_caption": combined_caption,
+        "enhanced_caption": enhanced_caption,
+        "image": image
+    }
+
+# Gradio interface
+with gr.Blocks() as demo:
+    gr.Markdown("# LeX-Enhancer & Lumina2 Demo")
+    gr.Markdown("Generate enhanced captions from simple image and text descriptions, then create images with Lumina2")
+    
+    with gr.Row():
+        with gr.Column():
+            image_caption = gr.Textbox(
+                lines=2,
+                label="Image Caption",
+                placeholder="Describe the visual content of the image",
+                value="A picture of a group of people gathered in front of a world map"
             )
-
-            seed = gr.Slider(
-                label="Seed",
-                minimum=0,
-                maximum=MAX_SEED,
-                step=1,
-                value=0,
+            text_caption = gr.Textbox(
+                lines=2,
+                label="Text Caption",
+                placeholder="Describe any text that should appear in the image",
+                value="\"Communicate\" in purple, \"Execute\" in yellow"
             )
-
-            randomize_seed = gr.Checkbox(label="Randomize seed", value=True)
-
-            with gr.Row():
-                width = gr.Slider(
-                    label="Width",
-                    minimum=256,
-                    maximum=MAX_IMAGE_SIZE,
-                    step=32,
-                    value=1024,  # Replace with defaults that work for your model
+            
+            with gr.Accordion("Advanced Settings", open=False):
+                seed = gr.Slider(
+                    minimum=0,
+                    maximum=100000,
+                    value=0,
+                    step=1,
+                    label="Seed (0 for random)"
                 )
-
-                height = gr.Slider(
-                    label="Height",
-                    minimum=256,
-                    maximum=MAX_IMAGE_SIZE,
-                    step=32,
-                    value=1024,  # Replace with defaults that work for your model
+                num_inference_steps = gr.Slider(
+                    minimum=20,
+                    maximum=100,
+                    value=80,
+                    step=1,
+                    label="Number of Inference Steps"
                 )
-
-            with gr.Row():
                 guidance_scale = gr.Slider(
-                    label="Guidance scale",
-                    minimum=0.0,
+                    minimum=1.0,
                     maximum=10.0,
+                    value=4.0,
                     step=0.1,
-                    value=0.0,  # Replace with defaults that work for your model
-                )
-
-                num_inference_steps = gr.Slider(
-                    label="Number of inference steps",
-                    minimum=1,
-                    maximum=50,
-                    step=1,
-                    value=2,  # Replace with defaults that work for your model
+                    label="Guidance Scale"
                 )
-
-        gr.Examples(examples=examples, inputs=[prompt])
-    gr.on(
-        triggers=[run_button.click, prompt.submit],
-        fn=infer,
-        inputs=[
-            prompt,
-            negative_prompt,
-            seed,
-            randomize_seed,
-            width,
-            height,
-            guidance_scale,
-            num_inference_steps,
-        ],
-        outputs=[result, seed],
+            
+            submit_btn = gr.Button("Generate", variant="primary")
+        
+        with gr.Column():
+            output_image = gr.Image(label="Generated Image")
+            combined_caption_box = gr.Textbox(
+                label="Combined Caption",
+                interactive=False
+            )
+            enhanced_caption_box = gr.Textbox(
+                label="Enhanced Caption",
+                interactive=False,
+                lines=5
+            )
+    
+    # Example prompts
+    examples = [
+        ["A modern office workspace", "\"Innovation\" in bold blue letters at the center"],
+        ["A beach sunset scene", "\"Relax\" in cursive white text in the corner"],
+        ["A futuristic city skyline", "\"The Future is Now\" in neon pink glowing letters"]
+    ]
+    gr.Examples(
+        examples=examples,
+        inputs=[image_caption, text_caption],
+        label="Example Inputs"
+    )
+    
+    submit_btn.click(
+        fn=run_pipeline,
+        inputs=[image_caption, text_caption, seed, num_inference_steps, guidance_scale],
+        outputs=[output_image, combined_caption_box, enhanced_caption_box]
     )
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch(server_name="0.0.0.0")