Update app.py

app.py

@@ -1,64 +1,84 @@
+
+from typing import List, Tuple, Dict, Generator
+from langchain.llms import OpenAI
 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)",
-        ),
+
+model_name = "gpt-4o"
+LLM = OpenAI(model_name=model_name, temperature=0.1)
+
+def create_history_messages(history: List[Tuple[str, str]]) -> List[dict]:
+    history_messages = [{"role": "user", "content": m[0]} for m in history]
+    history_messages.extend([{"role": "assistant", "content": m[1]} for m in history])
+    return history_messages
+
+def create_formatted_history(history_messages: List[dict]) -> List[Tuple[str, str]]:
+    formatted_history = []
+    user_messages = []
+    assistant_messages = []
+
+    for message in history_messages:
+        if message["role"] == "user":
+            user_messages.append(message["content"])
+        elif message["role"] == "assistant":
+            assistant_messages.append(message["content"])
+
+        if user_messages and assistant_messages:
+            formatted_history.append(
+                ("".join(user_messages), "".join(assistant_messages))
+            )
+            user_messages = []
+            assistant_messages = []
+
+    # append any remaining messages
+    if user_messages:
+        formatted_history.append(("".join(user_messages), None))
+    elif assistant_messages:
+        formatted_history.append((None, "".join(assistant_messages)))
+
+    return formatted_history
+
+def chat(
+    message: str, state: List[Dict[str, str]], client = LLM.client
+) -> Generator[Tuple[List[Tuple[str, str]], List[Dict[str, str]]], None, None]:
+    history_messages = state
+    if history_messages == None:
+        history_messages = []
+        history_messages.append({"role": "system", "content": "A helpful assistant."})
+
+    history_messages.append({"role": "user", "content": message})
+    # We have no content for the assistant's response yet but we will update this:
+    history_messages.append({"role": "assistant", "content": ""})
+
+    response_message = ""
+
+    chat_generator = client.create(
+        messages=history_messages, stream=True, model=model_name
+    )
+
+    for chunk in chat_generator:
+        if "choices" in chunk:
+            for choice in chunk["choices"]:
+                if "delta" in choice and "content" in choice["delta"]:
+                    new_token = choice["delta"]["content"]
+                    # Add the latest token:
+                    response_message += new_token
+                    # Update the assistant's response in our model:
+                    history_messages[-1]["content"] = response_message
+
+                if "finish_reason" in choice and choice["finish_reason"] == "stop":
+                    break
+        formatted_history = create_formatted_history(history_messages)
+        yield formatted_history, history_messages
+
+chatbot = gr.Chatbot(label="Chat").style(color_map=("yellow", "purple"))
+iface = gr.Interface(
+    fn=chat,
+    inputs=[
+        gr.Textbox(placeholder="Hello! How are you? etc.", label="Message"),
+        "state",
     ],
+    outputs=[chatbot, "state"],
+    allow_flagging="never",
 )
 
-
-if __name__ == "__main__":
-    demo.launch()
+iface.queue().launch()