| """ |
| LAB 1: What is RAG? — Side-by-Side Comparison |
| ================================================ |
| Plain LLM vs RAG-powered LLM using Abacus.AI RouteLLM. |
| Deploy on HuggingFace Spaces with Gradio. |
| """ |
|
|
| import os |
| import gradio as gr |
| from openai import OpenAI |
|
|
| |
| client = OpenAI( |
| base_url="https://routellm.abacus.ai/v1", |
| api_key=os.getenv("ABACUS_API_KEY"), |
| ) |
| MODEL = "gpt-4o-mini" |
|
|
| |
| KNOWLEDGE_BASE = """ |
| PavanCoaching AI Bootcamp starts on June 1, 2025. |
| The fee is ₹15,000 for the full course. |
| The course covers Python, ML, LLMs, RAG, and Agents. |
| Instructor: Pavan Kumar. Contact: pavan@pavancoaching.org |
| Batch size is limited to 30 students. |
| """ |
|
|
| def plain_llm(question: str) -> str: |
| """Ask the LLM WITHOUT any context.""" |
| try: |
| response = client.chat.completions.create( |
| model=MODEL, |
| messages=[{"role": "user", "content": question}], |
| ) |
| return response.choices[0].message.content |
| except Exception as e: |
| return f"❌ Error: {e}" |
|
|
| def rag_llm(question: str) -> str: |
| """Ask the LLM WITH retrieved context.""" |
| try: |
| response = client.chat.completions.create( |
| model=MODEL, |
| messages=[ |
| { |
| "role": "system", |
| "content": ( |
| "Answer ONLY using the context below.\n" |
| "If the answer is not in the context, say 'I don't know.'\n\n" |
| f"Context:\n{KNOWLEDGE_BASE}" |
| ), |
| }, |
| {"role": "user", "content": question}, |
| ], |
| temperature=0.1, |
| ) |
| return response.choices[0].message.content |
| except Exception as e: |
| return f"❌ Error: {e}" |
|
|
| with gr.Blocks(title="Lab 1: RAG vs Plain LLM", theme=gr.themes.Soft()) as demo: |
| gr.Markdown("## 🧪 Lab 1: Plain LLM vs RAG-Powered LLM") |
| question = gr.Textbox(label="Your Question", value="When does the AI Bootcamp start?") |
| btn = gr.Button("🔍 Compare", variant="primary") |
| with gr.Row(): |
| out_plain = gr.Textbox(label="❌ Plain LLM (no context)", lines=8) |
| out_rag = gr.Textbox(label="✅ RAG LLM (with context)", lines=8) |
| btn.click(fn=lambda q: (plain_llm(q), rag_llm(q)), inputs=question, outputs=[out_plain, out_rag]) |
|
|
| demo.launch() |
