Ship demo

app.py

@@ -1,63 +1,68 @@
+import re
+
 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
-"""
+from routellm.controller import Controller
+
+TEMPERATURE = 0.8
+THRESHOLD = 0.11593
+ROUTER = "mf"
+
+client = Controller(
+    routers=["mf"],
+    strong_model="gpt-4-1106-preview",
+    weak_model="anyscale/mistralai/Mixtral-8x7B-Instruct-v0.1",
+)
+
+
+def predict(message, history, threshold, temperature):
+    # Convert chat history to OpenAI format
+    history_openai_format = [
+        {"role": "system", "content": "You are a helpful AI assistant."}
+    ]
+    for human, assistant in history:
+        history_openai_format.append({"role": "user", "content": human})
+        history_openai_format.append(
+            {
+                "role": "assistant",
+                # Remove model name from response
+                "content": re.sub(r"^\[.*?\]\s*", "", assistant),
+            }
+        )
+    history_openai_format.append({"role": "user", "content": message})
+
+    # Create a chat completion request and send it to the API server
+    stream = client.chat.completions.create(
+        model=f"router-{ROUTER}-{threshold}",  # Model name to use
+        messages=history_openai_format,  # Chat history
+        temperature=temperature,  # Temperature for text generation
+        stream=True,  # Stream response
+    )
+    print(stream)
+
+    # Read and return generated text from response stream
+    partial_message = ""
+    for i, chunk in enumerate(stream):
+        print(chunk)
+        if i == 0:
+            model_prefix = f"[{chunk.model}]\n"
+            yield model_prefix
+            partial_message += model_prefix
+        partial_message += chunk.choices[0].delta.content or ""
+        yield partial_message
+
+
+# Create and launch a chat interface with Gradio
 demo = gr.ChatInterface(
-    respond,
+    predict,
     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(label="Threshold", minimum=0, maximum=1, value=THRESHOLD, step=0.01),
         gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
+            label="Temperature", minimum=0, maximum=1, value=TEMPERATURE, step=0.1
         ),
     ],
+    title="RouteLLM",
+    description="This is a demo of our matrix factorization router, calibrated so that approximately 50% of harder calls are routed to GPT-4, with remaining calls routed to Mixtral 8x7B.\nSee https://github.com/lm-sys/RouteLLM for details!",
 )
 
-
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()

requirements.txt

@@ -1 +1,2 @@
-huggingface_hub==0.22.2
+huggingface_hub==0.22.2
+routellm[serve]