Update app.py

app.py

@@ -1,64 +1,150 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
+from gradio import ChatMessage
 
+import json
+from openai import OpenAI
+from tools import tools, oitools
+from dotenv import load_dotenv
+import os
+load_dotenv(".env")
+HF_TOKEN = os.environ["HF_TOKEN"]
+BASE_URL = os.environ["BASE_URL"]
+
+SYSTEM_PROMPT_TEMPLATE = """
+You are an AI assistant designed to assist users with a hotel booking and information system. Your role is to provide detailed and accurate information about the hotel, including available accommodations, facilities, dining options, and reservation services. You can check room availability, assist with bookings, modify or cancel reservations, and answer general inquiries about the hotel.  
+
+Maintain clarity, conciseness, and relevance in your responses, ensuring a seamless user experience. Always respond in the same **language as the user’s query** to preserve their preferred language.
 """
-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")
 
+client = OpenAI(
+	    base_url=BASE_URL + "/v1", 
+	    api_key=HF_TOKEN
+    )
 
-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]})
+def complation(history, model, tools=None):
+    system_prompt = SYSTEM_PROMPT_TEMPLATE
+    messages = [{"role": "system", "content": system_prompt}]
+    for msg in history:
+        if type(msg) == dict:
+            msg = ChatMessage(**msg)
+        if msg.role == "assistant" and len(msg.options) > 0 and msg.options[0]["label"] == "tool_calls":
+            tools_calls = json.loads(msg.options[0]["value"])
+            messages.append({"role": "assistant", "tool_calls": tools_calls})
+            messages.append({"role": "tool", "content": msg.content})
+        else:
+            messages.append({"role": msg.role, "content": msg.content})
+    
+    if not tools:
+        return client.chat.completions.create(
+            model=model,
+            messages=messages,
+            stream=True,
+            max_tokens=1000,
+            temperature=0.4,
+            frequency_penalty=1,
+            # stop=["<|em_end|>"],
+            extra_body = {
+                "repetition_penalty": 1.1,
+            }
+        )
+    return client.chat.completions.create(
+            model=model,
+            messages=messages,
+            stream=True,
+            max_tokens=1000,
+            temperature=0.4,
+            tool_choice="auto",
+            tools=tools,
+            frequency_penalty=1,
+            # stop=["<|em_end|>"],
+            extra_body = {
+                "repetition_penalty": 1.1,
+            }
+        )
 
-    messages.append({"role": "user", "content": message})
+def respond(
+    message:any,
+    history:any,
+):
+    try:   
+        models = client.models.list()
+        model = models.data[0].id
+    except Exception as err:
+        gr.Warning("The model is initializing. Please wait; this may take 5 to 10 minutes ⏳.", duration=20)
+        raise err
 
     response = ""
+    arguments = ""
+    name = ""
+    history.append(
+        ChatMessage(
+            role="user",
+            content=message,
+        )
+    )
+    completion = complation(history=history,  tools=oitools, model=model)
+    appended = False
+    for chunk in completion:
+        if len(chunk.choices) > 0 and chunk.choices[0].delta.tool_calls and len(chunk.choices[0].delta.tool_calls) > 0 :
+            call = chunk.choices[0].delta.tool_calls[0]
+            if call.function.name:
+                name=call.function.name
+            if call.function.arguments:
+                arguments += call.function.arguments
 
-    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
+        elif chunk.choices[0].delta.content:
+            response += chunk.choices[0].delta.content
+            if not appended:
+                history.append(
+                   ChatMessage(
+                        role="assistant",
+                        content="",
+                    )
+                )
+                appended = True
+                
+            history[-1].content = response
+            yield history[-1]
+    
+    if not arguments:
+        arguments = "{}"
+    if name:
+        json_arguments = json.loads(arguments)
+        result = f"💥 Error using tool {name}, tools doesn't exists"
+        if tools.get(name):
+            result = str(tools[name].invoke(input=json_arguments))
+            result = json.dumps({name: result}, ensure_ascii=False)
+        history.append(
+                ChatMessage(
+                    role="assistant",
+                    content=result,
+                    metadata= {"title": f"🛠️ Used tool '{name}'"},
+                    options=[{"label":"tool_calls", "value": json.dumps([{"id": "call_FthC9qRpsL5kBpwwyw6c7j4k","function": {"arguments": arguments,"name": name},"type": "function"}])}]
+                )
+            )
+        yield history[-1]
 
+        completion = complation(history=history, tools=oitools, model=model)
+        result = ""
+        appended = False
+        for chunk in completion:
+            result += chunk.choices[0].delta.content
+            if not appended:
+                history.append(
+                   ChatMessage(
+                        role="assistant",
+                        content="",
+                    )
+                )
+                appended = True
+                
+            history[-1].content = result
+            yield history[-2:]
 
 """
 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()
+    demo = gr.ChatInterface(respond, type="messages")
+    demo.launch()