Spaces:
Sleeping
Sleeping
File size: 3,607 Bytes
6f7d7b4 4062a25 6f7d7b4 9178b0a 0f347bb 4062a25 0f347bb 9178b0a 9969cd9 1afe94e 9969cd9 1afe94e 0f347bb 9178b0a 0f347bb 9178b0a 6f7d7b4 9178b0a |
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 |
import gradio as gr
from huggingface_hub import InferenceClient
# Initialize a list to store the conversation history
conversation_history = []
# Function to interact with the model using the Inference API
def chat_with_model(user_input, hf_api_key):
global conversation_history
if not hf_api_key:
return "Error: Please provide your Hugging Face API key."
try:
# Initialize the InferenceClient with the provided API key
client = InferenceClient(api_key=hf_api_key)
# Add the user's message to the conversation history
conversation_history.append({"role": "user", "content": user_input})
# Define the system message (defining the assistant role)
system_message = {
"role": "system",
"content": "You are a code assistant that helps with code generation, debugging, and explanations."
}
# Add system message to the conversation history
if len(conversation_history) == 1: # Add system message only once
conversation_history.insert(0, system_message)
# Ensure the conversation history doesn't exceed token limits
if len(conversation_history) > 10: # Keep the last 10 messages
conversation_history = [system_message] + conversation_history[-10:]
# Create a stream for chat completions using the API
stream = client.chat.completions.create(
model="Qwen/Qwen2.5-Coder-32B-Instruct",
messages=conversation_history,
max_tokens=500,
stream=True
)
# Collect the generated response from the model
response = ""
for chunk in stream:
response += chunk.choices[0].delta.content
# Add the assistant's response to the conversation history
conversation_history.append({"role": "assistant", "content": response})
return response
except Exception as e:
return f"Error: {e}"
# Create the Gradio interface with a dark theme
with gr.Blocks(
css="""
body {
background-color: #121212;
color: #ffffff;
}
.gradio-container {
background-color: #1e1e1e;
}
.gradio-container input, .gradio-container textarea, .gradio-container button {
color: #ffffff;
background-color: #2e2e2e;
border: 1px solid #555555;
}
.gradio-container textarea::placeholder, .gradio-container input::placeholder {
color: #bbbbbb;
}
.gradio-container button {
background-color: #333333;
border: 1px solid #777777;
}
.gradio-container button:hover {
background-color: #444444;
}
#title, #description {
color: #ffffff !important;
}
"""
) as demo:
gr.Markdown("<h1 id='title'>Code Assistant with Qwen2.5-Coder</h1>")
gr.Markdown("<p id='description'>Ask me anything about coding! Enter your Hugging Face API key to start.</p>")
# Create the input and output interface
with gr.Row():
user_input = gr.Textbox(lines=5, placeholder="Ask me anything about coding...")
api_key = gr.Textbox(lines=1, placeholder="Enter your Hugging Face API key", type="password")
# Create the output display
output = gr.Textbox(label="Response")
# Button for submitting queries
submit_button = gr.Button("Submit")
submit_button.click(chat_with_model, inputs=[user_input, api_key], outputs=output)
# Launch the app
demo.launch() |