app.py

import gradio as gr
from llama_cpp import Llama
from huggingface_hub import hf_hub_download

# Global variable to track model loading status
model = None

def load_model():
    repo_id = "forestav/gguf_lora_model"
    model_file = "unsloth.F16.gguf"  
    
    local_path = hf_hub_download(
        repo_id=repo_id,
        filename=model_file
    )
    
    print(f"Loading model from: {local_path}")
    
    model = Llama(
        model_path=local_path,
        n_ctx=2048,
        n_threads=8
    )
    
    return model

def generate_career_response(message, history):
    if model is None:
        return "Model is still loading. Please wait..."
        
    enhanced_prompt = f"""As a career development advisor, help the user with their professional growth. 
    Consider:
    1. Skill development opportunities
    2. Industry trends
    3. Practical next steps
    4. Resources and learning paths
    
    User Query: {message}
    
    Provide a structured response with actionable advice."""

    response = model.create_chat_completion(
        messages=[
            {"role": "system", "content": "You are a professional career advisor focused on providing practical, actionable guidance for career development."},
            {"role": "user", "content": enhanced_prompt}
        ],
        max_tokens=512,
        temperature=0.7,
        top_p=0.95,
    )
    
    return response['choices'][0]['message']['content']

# Create the interface first
demo = gr.ChatInterface(
    fn=generate_career_response,
    title="Career Growth Navigator 🚀",
    description="""Your AI career development partner. Ask about:
    • Skill development paths
    • Career transition strategies
    • Industry trends and opportunities
    • Resume and interview preparation
    • Professional networking advice
    • Work-life balance
    Let's shape your professional future together!""",
    examples=[
        "I'm a software developer wanting to transition into AI/ML. What skills should I focus on?",
        "How can I improve my leadership skills in my current role?",
        "What are the key trends in digital marketing I should be aware of?",
        "I want to start freelancing in web development. Where should I begin?",
        "How can I negotiate a promotion in my current position?"
    ]
)

# Create loading interface
with gr.Blocks() as loading_demo:
    gr.Markdown("# Loading Career Growth Navigator 🚀")
    with gr.Row():
        loading_msg = gr.Markdown("⏳ The model is currently loading. Please wait...")

if __name__ == "__main__":
    # Start with loading interface
    loading_demo.queue()
    loading_demo.launch(
        server_name="0.0.0.0",
        server_port=7860,
        share=False,
        prevent_thread_lock=True
    )
    
    # Load the model
    print("Starting model loading...")
    model = load_model()
    print("Model loaded successfully!")
    
    # Close loading interface and launch main interface
    loading_demo.close()
    demo.queue()
    demo.launch(
        server_name="0.0.0.0",
        server_port=7860,
        share=False
    )