Update app.py

app.py

@@ -1,7 +1,7 @@
 from fastapi import FastAPI
 from pydantic import BaseModel
 from transformers import AutoTokenizer, AutoModelForCausalLM
-import torch
+import torch, gradio as gr
 
 MODEL_ID = "Gopu-poss/gopu-agent-2k-fdf"
 
@@ -23,3 +23,25 @@ def infer(data: Input):
     outputs = model.generate(**inputs, max_new_tokens=200, temperature=0.7, top_p=0.9)
     text = tokenizer.decode(outputs[0], skip_special_tokens=True)
     return {"generated_text": text}
+
+# Gradio multimodal
+def gopu_chat(prompt, image=None, audio=None, video=None):
+    # Pour l’instant on ne traite que le texte
+    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
+    outputs = model.generate(**inputs, max_new_tokens=200, temperature=0.7, top_p=0.9)
+    return tokenizer.decode(outputs[0], skip_special_tokens=True)
+
+demo = gr.Interface(
+    fn=gopu_chat,
+    inputs=[
+        gr.Textbox(label="Texte"),
+        gr.Image(label="Image", type="filepath", optional=True),
+        gr.Audio(label="Voix", type="filepath", optional=True),
+        gr.Video(label="Vidéo", type="filepath", optional=True)
+    ],
+    outputs="text",
+    title="GopuOS Agentic Endpoint",
+    description="Texte, image, voix, vidéo, recherche web (à venir)."
+)
+
+app = gr.mount_gradio_app(app, demo, path="/gradio")