Create app.py

app.py

@@ -0,0 +1,56 @@
+import gradio as gr
+from gradio_client import Client, handle_file
+import requests
+from PIL import Image
+import io
+
+def predict(image, question, seed, top_p, temperature):
+    # Inicializa o cliente do Gradio
+    client = Client("deepseek-ai/Janus-Pro-7B")
+    
+    # Prepara a imagem para envio
+    if image.startswith('http'):
+        response = requests.get(image)
+        img_path = handle_file(io.BytesIO(response.content))
+    else:
+        img_path = handle_file(image)
+    
+    # Faz a predição
+    result = client.predict(
+        image=img_path,
+        question=question,
+        seed=seed,
+        top_p=top_p,
+        temperature=temperature,
+        api_name="/multimodal_understanding"
+    )
+    
+    return result
+
+# Componentes da interface
+image_input = gr.Image(label="Upload Image", type="filepath")
+question_input = gr.Textbox(label="Question", placeholder="Ask something about the image...")
+seed_slider = gr.Slider(0, 100, value=42, label="Seed")
+top_p_slider = gr.Slider(0, 1, value=0.95, label="Top-p")
+temp_slider = gr.Slider(0, 1, value=0.1, label="Temperature")
+
+# Cria a interface
+demo = gr.Interface(
+    fn=predict,
+    inputs=[
+        image_input,
+        question_input,
+        seed_slider,
+        top_p_slider,
+        temp_slider
+    ],
+    outputs=gr.Textbox(label="Answer"),
+    title="Janus-Pro-7B Multimodal Demo",
+    description="Ask questions about images using the Janus-Pro-7B model",
+    examples=[
+        ["https://raw.githubusercontent.com/gradio-app/gradio/main/test/test_files/bus.png", "What's in this image?", 42, 0.95, 0.1]
+    ]
+)
+
+if __name__ == "__main__":
+    demo.launch()