upload app.py

app.py

@@ -1,64 +1,81 @@
-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
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
-
-
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
-
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
-
-    messages.append({"role": "user", "content": message})
-
-    response = ""
-
-    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
-
-        response += token
-        yield response
-
-
-"""
-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)",
-        ),
-    ],
-)
-
-
-if __name__ == "__main__":
-    demo.launch()
+from coordinator_agent import coordinator_agent
+from coordinator_agent_langgraph import  AgentState, coordinator_graph
+from dotenv import load_dotenv
+from langchain_core.messages import convert_to_messages
+import gradio as gr
+from langchain_google_genai import ChatGoogleGenerativeAI 
+
+load_dotenv()
+
+
+
+# demo = gr.Interface(
+#     fn=lambda query: "\n\n".join(
+#         f"🔹 {msg.name if hasattr(msg, 'name') else 'Agent'}:\n{msg.content}"
+#         for msg in coordinator_agent.invoke({"messages": [{"role": "user", "content": query}]}).get("messages", [])
+#         if hasattr(msg, "content") and msg.content
+#     ),
+#     inputs=gr.Textbox(label="📝 Frag etwas zum Markt", placeholder="z.B. Was gibt es Neues bei NVIDIA?"),
+#     outputs=gr.Textbox(label="🤖 Antwort der Agenten"),
+#     title="📊 Multimodaler Markt-Analyst (Gradio)",
+#     description="Ein intelligentes System zur Analyse von Marktinformationen mit mehreren Agenten.",
+# )
+
+#   Langchain Version
+def run_supervisor_full(query):
+    result = coordinator_agent.invoke({"messages": [{"role": "user", "content": query}]})
+    history = result.get("messages", [])
+
+    chunks = []
+    for msg in history:
+        name = getattr(msg, "name", "Agent")
+        if hasattr(msg, "content") and msg.content:
+            content = msg.content
+        elif hasattr(msg, "tool_call_id"):
+            content = f"[→ Übergabe an Tool: {msg.tool}]"
+        else:
+            continue
+        chunks.append(f"🔹 {name}:\n{content}")
+
+    return "\n\n".join(chunks)
+# demo = gr.Interface(
+#     fn=lambda query: "\n\n".join(
+#         f"🔹 {msg.name if hasattr(msg, 'name') else 'Agent'}:\n"
+#         f"{msg.content}"
+#         + (f"\n\n🔗 Quelle: {msg.metadata.get('source')}" if hasattr(msg, "metadata") and "source" in msg.metadata else "")
+#         for msg in coordinator_agent.invoke({"messages": [{"role": "user", "content": query}]}).get("messages", [])
+#         if hasattr(msg, "content") and msg.content
+#     ),
+#     inputs=gr.Textbox(label="📝 Frag etwas zum Markt", placeholder="z.B. Was gibt es Neues bei NVIDIA?"),
+#     outputs=gr.Textbox(label="🤖 Antwort der Agenten"),
+#     title="📊 Multimodaler Markt-Analyst (Gradio)",
+#     description="Ein intelligentes System zur Analyse von Marktinformationen mit mehreren Agenten.",
+# )
+
+demo = gr.Interface(
+    fn=run_supervisor_full,
+    inputs=gr.Textbox(label="📝 Marktfrage", placeholder="z. B. Wie war die Performance von NVIDIA 2023?"),
+    outputs=gr.Textbox(label="🤖 Antwortverlauf"),
+    title="🧠 Koordinator-Agent mit LangGraph",
+    description="Automatisches Routing zu spezialisierten Agenten mit vollständigem Nachrichtenverlauf."
+)
+
+
+# Langgraph Version
+# def run_coordinator(user_input):
+#     result = coordinator_graph.invoke(AgentState({"input": user_input}))
+#     return result.get("response", "Keine Antwort erhalten.")
+# demo = gr.Interface(
+#     fn=run_coordinator,
+#     inputs=gr.Textbox(label="📝 Deine Marktfrage", placeholder="z. B. Wie war NVIDIAs Ergebnis 2023?"),
+#     outputs=gr.Textbox(label="🤖 Antwort vom System"),
+#     title="📊 Multimodaler Markt-Analyst (LangGraph)",
+#     description="Ein intelligentes System mit mehreren spezialisierten Agenten (Finanz, Analyse, Web)."
+# )
+
+if __name__ == "__main__":
+    # user_question = "Wie war die Performance von Apple im Jahr 2023?"
+    # print("▶️ Direktaufruf:")
+    # print(run_coordinator(user_question))
+    demo.launch()
+