Update app.py

app.py

@@ -1,64 +1,128 @@
-import gradio as gr
-from huggingface_hub import InferenceClient
-
 """
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
+Gradio Space: Text → Image using FLUX.1 (Hugging Face Inference API)
+Attractive interface with custom styling and footer label: "designed by Mehak Mazhar"
+
+How to use:
+1. Install dependencies: pip install gradio requests pillow
+2. Get a Hugging Face API token (if you want to use the hosted FLUX.1 models) and either set it as an env var HF_TOKEN or paste it into the 'HF Token' field in the UI.
+3. Run: python gradio_flux_text2img.py
+
+Notes: this script calls the Hugging Face Inference API for the model 'black-forest-labs/FLUX.1-schnell' by default.
+You can change the MODEL variable to any compatible image generation model hosted on Hugging Face or point to your own inference server.
+
 """
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
 
+import os
+import io
+import base64
+import random
+import requests
+from PIL import Image
+import gradio as gr
 
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
+# --- Configuration ---
+MODEL = os.environ.get("FLUX_MODEL", "black-forest-labs/FLUX.1-schnell")
+HF_API_URL = f"https://api-inference.huggingface.co/models/{MODEL}"
 
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
+# A small helper to call the Hugging Face Inference API (image generation)
+def call_hf_image_api(prompt, token, width, height, guidance_scale, steps, seed, negative_prompt=None):
+    headers = {"Authorization": f"Bearer {token}"} if token else {}
+    payload = {
+        "inputs": prompt,
+        "options": {"wait_for_model": True},
+        "parameters": {
+            "width": int(width),
+            "height": int(height),
+            "guidance_scale": float(guidance_scale),
+            "num_inference_steps": int(steps),
+            "seed": None if seed is None else int(seed),
+        }
+    }
+    if negative_prompt:
+        payload["parameters"]["negative_prompt"] = negative_prompt
 
-    messages.append({"role": "user", "content": message})
+    # Many HF image models return binary image bytes directly. Some return JSON with base64.
+    resp = requests.post(HF_API_URL, headers=headers, json=payload, stream=True, timeout=120)
+    resp.raise_for_status()
 
-    response = ""
+    content_type = resp.headers.get("content-type", "")
+    if "application/json" in content_type:
+        data = resp.json()
+        # attempt to find a base64 image in JSON response
+        # common key patterns: 'image', 'images', 'generated_images'
+        if isinstance(data, dict):
+            for k in ("image", "images", "generated_images", "artifacts"):
+                if k in data:
+                    imgs = data[k]
+                    if isinstance(imgs, list) and len(imgs) > 0:
+                        b64 = imgs[0].get("data") if isinstance(imgs[0], dict) else imgs[0]
+                        if isinstance(b64, str):
+                            return Image.open(io.BytesIO(base64.b64decode(b64)))
+        # fallback: try to find base64 strings in the JSON
+        for v in data.values():
+            if isinstance(v, str) and v.strip().startswith("iVBOR"):  # PNG base64 signature
+                return Image.open(io.BytesIO(base64.b64decode(v)))
+        raise ValueError("Could not parse image from JSON response")
+    else:
+        # assume raw image bytes
+        return Image.open(io.BytesIO(resp.content))
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
+# --- Gradio UI ---
+css = r'''
+body { background: linear-gradient(135deg, #fff7e6 0%, #fffaf0 50%, #fff7fd 100%); }
+.gradio-container { font-family: 'Inter', system-ui, -apple-system, 'Segoe UI', Roboto, 'Helvetica Neue', Arial; }
+.header { display:flex; align-items:center; gap:16px; }
+.logo { width:64px; height:64px; border-radius:14px; box-shadow: 0 6px 18px rgba(0,0,0,0.08); }
+.card { background: rgba(255,255,255,0.9); border-radius:16px; padding:18px; box-shadow: 0 12px 30px rgba(0,0,0,0.06); }
+.footer { text-align:center; font-size:12px; color:#555; margin-top:12px; }
+.footer strong { color:#333; }
+.generator-btn { border-radius: 12px; padding:10px 18px; }
+'''
 
-        response += token
-        yield response
+with gr.Blocks(css=css, title="Flux Text→Image — designed by Mehak Mazhar") as demo:
+    with gr.Row(elem_id="top-row"):
+        with gr.Column(scale=1):
+            gr.HTML("<div class='header'><img class='logo' src='https://raw.githubusercontent.com/black-forest-labs/flux/main/logo.png' alt='Flux logo' onerror='this.style.display=\'none\'' + "> <div><h2 style='margin:0'>FLUX.1 Text → Image</h2><p style='margin:0;color:#555;'>Generate high-quality images from text (Hugging Face inference API)</p></div></div>")
 
+    with gr.Row():
+        with gr.Column(scale=1, min_width=360):
+            prompt = gr.Textbox(label="Prompt", placeholder="A serene mountain lake at sunset, digital painting, ultra-detailed", rows=4)
+            negative = gr.Textbox(label="Negative prompt (optional)", placeholder="blurry, lowres, text, watermark", rows=2)
+            hf_token = gr.Textbox(label="Hugging Face API Token (optional)", placeholder="Paste your HF token here or set HF_TOKEN env var", type="password")
+            with gr.Row():
+                width = gr.Dropdown(choices=[256,384,512,768,1024], value=512, label="Width")
+                height = gr.Dropdown(choices=[256,384,512,768,1024], value=512, label="Height")
+            with gr.Row():
+                steps = gr.Slider(10, 150, value=28, step=1, label="Steps")
+                guidance = gr.Slider(1.0, 30.0, value=7.5, step=0.1, label="Guidance scale")
+            with gr.Row():
+                seed = gr.Number(value=None, precision=0, label="Seed (leave blank for random)")
+            gen_btn = gr.Button("Generate", elem_classes="generator-btn")
+            info = gr.Markdown("**Tip:** Use vivid, descriptive prompts. Try styles like `cinematic lighting`, `digital oil painting`, or `ultra-detailed`.")
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
-)
+        with gr.Column(scale=1, min_width=360):
+            gallery = gr.Gallery(label="Generated images", show_label=True, elem_id="gallery").style(grid=[2], height="640px")
+            out_log = gr.Textbox(label="Status / Debug log", lines=4, interactive=False)
+
+    # Footer with the requested label
+    gr.HTML("<div class='footer'>\n<p><strong>designed by Mehak Mazhar</strong></p>\n</div>")
+
+    def generate_image(prompt_text, negative_text, hf_token_text, width_v, height_v, steps_v, guidance_v, seed_v):
+        token = hf_token_text.strip() or os.environ.get("HF_TOKEN")
+        if not token:
+            return None, "ERROR: No Hugging Face token provided. Set HF_TOKEN or paste it into the UI."
+        try:
+            if seed_v in (None, "", 0):
+                seed_val = random.randint(0, 2**31 - 1)
+            else:
+                seed_val = int(seed_v)
+            img = call_hf_image_api(prompt_text, token, width_v, height_v, guidance_v, steps_v, seed_val, negative_prompt=negative_text)
+            return [img], f"OK — seed={seed_val}, model={MODEL}"
+        except Exception as e:
+            return None, f"API error: {e}"
 
+    gen_btn.click(fn=generate_image, inputs=[prompt, negative, hf_token, width, height, steps, guidance, seed], outputs=[gallery, out_log])
 
+# Run the app when executed directly
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch(server_name="0.0.0.0", share=False)