Update app.py

Follow example from flux--dev

app.py

@@ -1,19 +1,26 @@
-from PIL import Image
 import gradio as gr
 import torch
-from diffusers import FluxPipeline
 from groq import Groq  # Import the Groq library
 from cryptography.fernet import Fernet
 from huggingface_hub import login
-# from transformers import CLIPTokenizer
 import os
 os.environ['HF_HUB_DOWNLOAD_TIMEOUT'] = '120'
+import numpy as np
+import random
+import spaces
+from diffusers import  DiffusionPipeline, FlowMatchEulerDiscreteScheduler, AutoencoderTiny, AutoencoderKL
+from transformers import CLIPTextModel, CLIPTokenizer,T5EncoderModel, T5TokenizerFast
+from live_preview_helpers import calculate_shift, retrieve_timesteps, flux_pipe_call_that_returns_an_iterable_of_images
+os.environ['HF_HUB_DOWNLOAD_TIMEOUT'] = '120'
+dtype = torch.bfloat16
+device = "cuda" if torch.cuda.is_available() else "cpu"
 def get_hf_token(encrypted_token):
     # Retrieve the decryption key from an environment variable
     key = "K4FlQbffvTcDxT2FIhrOPV1eue6ia45FFR3kqp2hHbM="
     if not key:
         raise ValueError("Missing decryption key! Set the DECRYPTION_KEY environment variable.")
-
+    
+    # Convert key from string to bytes if necessary
     if isinstance(key, str):
         key = key.encode()
     
@@ -25,10 +32,19 @@ groq_client = Groq(api_key="gsk_0Rj7v0ZeHyFEpdwUMBuWWGdyb3FYGUesOkfhi7Gqba9rDXwI
 decrypted_token = get_hf_token("gAAAAABn3GfShExoJd50nau3B5ZJNiQ9dRD1ACO3XXMwVaIQMkmi59cL-MKGr6SYnsB0E2gGITJG2j29Ar9yjaZP-EC6hHsCBmwKSj4aFtTor9_n0_NdMBv1GtlxZRmwnQwriB-Xr94e")
 login(token=decrypted_token)
 
+taef1 = AutoencoderTiny.from_pretrained("madebyollin/taef1", torch_dtype=dtype).to(device)
+good_vae = AutoencoderKL.from_pretrained("black-forest-labs/FLUX.1-dev", subfolder="vae", torch_dtype=dtype).to(device)
+pipe = DiffusionPipeline.from_pretrained("black-forest-labs/FLUX.1-dev", torch_dtype=dtype, vae=taef1).to(device)
+torch.cuda.empty_cache()
 
-pipe = FluxPipeline.from_pretrained("black-forest-labs/FLUX.1-schnell", torch_dtype=torch.float32)
-pipe.enable_model_cpu_offload() #save some VRAM by offloading the model to CPU. Remove this if you have enough GPU power
+MAX_SEED = np.iinfo(np.int32).max
+MAX_IMAGE_SIZE = 2048
 
+pipe.flux_pipe_call_that_returns_an_iterable_of_images = flux_pipe_call_that_returns_an_iterable_of_images.__get__(pipe)
+
+decrypted_token = get_hf_token("gAAAAABn3GfShExoJd50nau3B5ZJNiQ9dRD1ACO3XXMwVaIQMkmi59cL-MKGr6SYnsB0E2gGITJG2j29Ar9yjaZP-EC6hHsCBmwKSj4aFtTor9_n0_NdMBv1GtlxZRmwnQwriB-Xr94e")
+login(token=decrypted_token)
+groq_client = Groq(api_key="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")
 
 def enhance_prompt(user_prompt):
     """Enhances the given prompt using Groq and returns the refined prompt."""
@@ -63,67 +79,117 @@ def enhance_prompt(user_prompt):
     except Exception as e:
         enhanced = f"Error enhancing prompt: {str(e)}"
     return enhanced
-# tokenizer = CLIPTokenizer.from_pretrained("openai/clip-vit-large-patch14")
+# ---------- End Prompt Enhancement Setup ----------
 
-# def truncate_prompt(prompt, max_length=77):
-#     # Tokenize without adding special tokens
-#     tokens = tokenizer(prompt, add_special_tokens=False)["input_ids"]
-#     if len(tokens) > max_length:
-#         tokens = tokens[:max_length]
-#         # Decode tokens back to a string (skipping special tokens)
-#         prompt = tokenizer.decode(tokens, skip_special_tokens=True)
-#     return prompt
+@spaces.GPU(duration=75)
+def infer(prompt, seed=42, randomize_seed=False, width=1024, height=1024, guidance_scale=3.5, num_inference_steps=28, progress=gr.Progress(track_tqdm=True)):
+    # First, refine the prompt using the enhancer
+    enhanced_prompt = enhance_prompt(prompt)
+    print("Enhanced prompt:", enhanced_prompt)  # You might remove or modify this logging
 
-def generate_image(prompt):
-    """Generates an image using the refined prompt."""
-    # truncated_prompt = truncate_prompt(prompt)
-    try:
-        image = pipe(
-        prompt,
-        height=1024,
-        width=1024,
-        guidance_scale=3.5,
-        num_inference_steps=50,
-        max_sequence_length=512,
-        generator=torch.Generator("cpu").manual_seed(0)
-).images[0]
-    except Exception as e:
-        # Optionally, handle errors (you can also return a default error image)
-        image = None
-        print(e)
-    return image
+    if randomize_seed:
+        seed = random.randint(0, MAX_SEED)
+    generator = torch.Generator().manual_seed(seed)
+    
+    # Use the enhanced prompt for image generation
+    for img in pipe.flux_pipe_call_that_returns_an_iterable_of_images(
+            prompt=enhanced_prompt,
+            guidance_scale=guidance_scale,
+            num_inference_steps=num_inference_steps,
+            width=width,
+            height=height,
+            generator=generator,
+            output_type="pil",
+            good_vae=good_vae,
+        ):
+            yield img, seed
 
-with gr.Blocks(css=".gradio-container {background-color: #f9f9f9; padding: 20px;}") as demo:
-    gr.Markdown("# 2-Step Image Generator")
-    gr.Markdown(
-        "### Step 1: Prompt Enhancement\n"
-        "Enter your original prompt below and click **Enhance Prompt**. "
-        "The system will generate a detailed version of your prompt. You can modify the enhanced prompt before generating the image."
-    )
+css = """
+#col-container {
+    margin: 0 auto;
+    max-width: 520px;
+}
+"""
+
+with gr.Blocks(css=css) as demo:
     
-    with gr.Row():
-        original_prompt = gr.Textbox(
-            label="Your Original Prompt",
-            placeholder="Describe your idea here...",
-            lines=3
+    with gr.Column(elem_id="col-container"):
+        gr.Markdown(f"""# FLUX.1 [dev] with Prompt Enhancement  
+Use the enhanced prompt (generated by Groq) for a refined image generation experience.
+        """)
+        
+        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)
+        
+        result = gr.Image(label="Result", show_label=False)
+        
+        with gr.Accordion("Advanced Settings", open=False):
+            seed = gr.Slider(
+                label="Seed",
+                minimum=0,
+                maximum=MAX_SEED,
+                step=1,
+                value=0,
+            )
+            
+            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,
+                )
+                height = gr.Slider(
+                    label="Height",
+                    minimum=256,
+                    maximum=MAX_IMAGE_SIZE,
+                    step=32,
+                    value=1024,
+                )
+            
+            with gr.Row():
+                guidance_scale = gr.Slider(
+                    label="Guidance Scale",
+                    minimum=1,
+                    maximum=15,
+                    step=0.1,
+                    value=3.5,
+                )
+                num_inference_steps = gr.Slider(
+                    label="Number of inference steps",
+                    minimum=1,
+                    maximum=50,
+                    step=1,
+                    value=28,
+                )
+        
+        gr.Examples(
+            examples=[
+                "a tiny astronaut hatching from an egg on the moon",
+                "a cat holding a sign that says hello world",
+                "an anime illustration of a wiener schnitzel",
+            ],
+            fn=infer,
+            inputs=[prompt],
+            outputs=[result, seed],
+            cache_examples="lazy"
         )
-        enhance_button = gr.Button("Enhance Prompt")
-    
-    enhanced_prompt_box = gr.Textbox(
-        label="Enhanced Prompt (Editable)",
-        placeholder="The enhanced prompt will appear here...",
-        lines=3
+
+    gr.on(
+        triggers=[run_button.click, prompt.submit],
+        fn=infer,
+        inputs=[prompt, seed, randomize_seed, width, height, guidance_scale, num_inference_steps],
+        outputs=[result, seed]
     )
-    
-    enhance_button.click(fn=enhance_prompt, inputs=original_prompt, outputs=enhanced_prompt_box)
-    
-    gr.Markdown("### Step 2: Image Generation\n"
-                "Review and modify the enhanced prompt if necessary. Once you're ready, click **Generate Image** to create your image.")
-    
-    generate_button = gr.Button("Generate Image")
-    image_output = gr.Image(label="Generated Image")
-    
-    generate_button.click(fn=generate_image, inputs=enhanced_prompt_box, outputs=image_output)
 
-if __name__ == "__main__":
-    demo.launch(share=True)
+demo.launch(share=True)