Spaces:
Sleeping
Sleeping
File size: 3,209 Bytes
f846bf2 5e0a7d2 f846bf2 064bf6d 5e0a7d2 f846bf2 064bf6d f846bf2 5e0a7d2 f846bf2 064bf6d 5e0a7d2 f846bf2 5e0a7d2 f846bf2 5e0a7d2 f846bf2 064bf6d f846bf2 5e0a7d2 f846bf2 5e0a7d2 f846bf2 5e0a7d2 f846bf2 5e0a7d2 a7e4213 |
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 |
# app.py
import gradio as gr
from src.agent import Agent
from src.create_database import load_and_process_dataset # Import from create_database.py
import os
import uuid
import requests
import logging
import subprocess
# Configure logging
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
# Function to install requirements
def install_requirements():
try:
subprocess.check_call([os.sys.executable, '-m', 'pip', 'install', '-r', 'requirements.txt'])
logging.info("Requirements installed successfully.")
except subprocess.CalledProcessError as e:
logging.error(f"Failed to install requirements: {e}")
# Install requirements
install_requirements()
# Create the directory if it doesn't exist
local_dir = "models"
os.makedirs(local_dir, exist_ok=True)
# Specify the filename for the model
filename = "unsloth.Q4_K_M.gguf"
model_path = os.path.join(local_dir, filename)
# Function to download the model file
def download_model(repo_id, filename, save_path):
# Construct the URL for the model file
url = f"https://huggingface.co/{repo_id}/resolve/main/{filename}"
# Download the model file
response = requests.get(url)
if response.status_code == 200:
with open(save_path, 'wb') as f:
f.write(response.content)
print(f"Model downloaded to {save_path}")
else:
print(f"Failed to download model: {response.status_code}")
# Download the model if it doesn't exist
if not os.path.exists(model_path):
download_model("PurpleAILAB/Llama3.2-3B-uncensored-SQLi-Q4_K_M-GGUF", filename)
def respond(
message,
history: list[tuple[str, str]],
system_message,
):
model_path = "models/unsloth.Q4_K_M.gguf" # Path to the downloaded model
db_path = "agent.db"
system_prompt = system_message
# Check if the database exists, if not, initialize it
if not os.path.exists(db_path):
data_update_path = "data-update.txt"
keyword_dir = "keyword" # Updated keyword directory
load_and_process_dataset(data_update_path, keyword_dir, db_path)
# Load the model with the maximum context length and control the maximum tokens in the response
llm = Llama(
model_path=model_path,
n_ctx=5000, # Set the maximum context length
max_tokens=512 # Control the maximum number of tokens generated in the response
)
agent = Agent(llm, db_path, system_prompt)
user_id = str(uuid.uuid4()) # Generate a unique user ID for each session
response = agent.process_query(user_id, message)
return 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="Vous êtes l'assistant intelligent de Les Chronique MTC. Votre rôle est d'aider les visiteurs en expliquant le contenu des Chroniques, Flash Infos et Chronique-FAQ de Michel Thomas. Utilisez le contexte fourni pour améliorer vos réponses et veillez à ce qu'elles soient précises et pertinentes.", label="System message"),
],
)
if __name__ == "__main__":
demo.launch()
|