File size: 2,793 Bytes
c67e035
 
 
ba6c3d0
 
c67e035
 
ba6c3d0
c67e035
ba6c3d0
 
 
 
c67e035
ba6c3d0
 
 
 
 
 
 
 
 
 
c67e035
ba6c3d0
 
 
 
 
 
 
 
 
 
 
 
c67e035
ba6c3d0
c67e035
 
ba6c3d0
c67e035
 
ba6c3d0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c67e035
ba6c3d0
c67e035
 
 
 
 
 
 
 
 
 
 
 
ba6c3d0
 
c67e035
 
ba6c3d0
c67e035
ba6c3d0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
import gradio as gr
from huggingface_hub import InferenceClient

# Initialize the InferenceClient with the model ID from Hugging Face
client = InferenceClient(model="HuggingFaceH4/zephyr-7b-beta")

def respond(
    message: str,
    history: list[tuple[str, str]],
    system_message: str,
    max_tokens: int,
    temperature: float,
    top_p: float,
):
    """
    Generates a response from the AI model based on the user's message and chat history.
    
    Args:
        message (str): The user's input message.
        history (list): A list of tuples representing the conversation history (user, assistant).
        system_message (str): A system-level message guiding the AI's behavior.
        max_tokens (int): The maximum number of tokens for the output.
        temperature (float): Sampling temperature for controlling the randomness.
        top_p (float): Top-p (nucleus sampling) for controlling diversity.

    Yields:
        str: The AI's response as it is generated.
    """
    
    # Prepare the conversation history for the API call
    messages = [{"role": "system", "content": system_message}]
    
    for user_input, assistant_response in history:
        if user_input:
            messages.append({"role": "user", "content": user_input})
        if assistant_response:
            messages.append({"role": "assistant", "content": assistant_response})

    # Add the latest user message to the conversation
    messages.append({"role": "user", "content": message})

    # Initialize an empty response
    response = ""

    try:
        # Generate a response from the model with streaming
        for message in client.chat_completion(
            messages=messages,
            max_tokens=max_tokens,
            stream=True,
            temperature=temperature,
            top_p=top_p,
        ):
            token = message.choices[0].delta.content
            response += token
            yield response

    except Exception as e:
        yield f"An error occurred: {str(e)}"


# Define the ChatInterface with additional input components for user customization
demo = gr.ChatInterface(
    fn=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)",
        ),
    ],
    title="Chatbot Interface",
    description="A customizable chatbot interface using Hugging Face's Inference API.",
)

# Launch the Gradio interface
if __name__ == "__main__":
    demo.launch()