File size: 4,230 Bytes
179f492 439f105 179f492 439f105 179f492 439f105 179f492 d856f89 439f105 d856f89 ca99935 439f105 ca99935 d856f89 ca99935 439f105 179f492 439f105 179f492 |
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 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 |
import gradio as gr
from huggingface_hub import InferenceClient
# Define the InferenceClient for different models
client_chatgpt = InferenceClient("openai/gpt-3.5-turbo") # Example for ChatGPT model
client_llama = InferenceClient("meta-llama/Llama-3.2-3B-Instruct") # Llama model
client_claude = InferenceClient("anthropic/claude-1") # Claude model (adjust with correct model path)
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_llama.chat_completion( # Defaulting to Llama, update dynamically later
messages,
max_tokens=max_tokens,
stream=True,
temperature=temperature,
top_p=top_p,
):
token = message.choices[0].delta.content
response += token
yield response
# Function to handle button clicks for different models
def on_button_click(model_name, message, history, system_message, max_tokens, temperature, top_p):
# Choose the client based on the selected model
if model_name == "Chatgpt":
client = client_chatgpt
elif model_name == "Llama":
client = client_llama
elif model_name == "Claude":
client = client_claude
else:
return "Unknown model selected."
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 = ""
# Call the selected model for completion
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
# CSS for styling the interface
css = """
body {
background-color: #06688E; /* Dark background */
color: white; /* Text color for better visibility */
}
.gr-button {
background-color: #42B3CE !important; /* White button color */
color: black !important; /* Black text for contrast */
border: none !important;
padding: 8px 16px !important;
border-radius: 5px !important;
}
.gr-button:hover {
background-color: #e0e0e0 !important; /* Slightly lighter button on hover */
}
.gr-slider-container {
color: white !important; /* Slider labels in white */
}
"""
# Interface using Blocks context
with gr.Blocks() as demo:
# Add all your components here, including buttons
system_message = gr.Textbox(value="You are a virtual health assistant...", label="System message", visible=False)
message_input = gr.Textbox(label="User message")
history = gr.State([]) # Keep the history of interactions
max_tokens_slider = gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens")
temperature_slider = gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature")
top_p_slider = gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)")
# Buttons to select the model
gr.Button("Chatgpt").click(on_button_click, inputs=[message_input, history, system_message, max_tokens_slider, temperature_slider, top_p_slider], outputs="text")
gr.Button("Llama").click(on_button_click, inputs=[message_input, history, system_message, max_tokens_slider, temperature_slider, top_p_slider], outputs="text")
gr.Button("Claude").click(on_button_click, inputs=[message_input, history, system_message, max_tokens_slider, temperature_slider, top_p_slider], outputs="text")
# Optional: customize your layout with CSS if needed
demo.css = css
# Launch the app
if __name__ == "__main__":
demo.launch(share=True)
|