Question 1

How to set up the Hands-on LLMs course on my local machine?

Accepted Answer

You need to install dependencies per module using Poetry, set up external service accounts (Alpaca, Qdrant, etc.), and configure cloud credentials; each module has its own README with specific steps, but the process is complex and requires careful environment variable management.

Question 2

Hands-on LLMs vs Andrew Ng's courses for learning LLM practical skills?

Accepted Answer

Hands-on LLMs focuses on a complete production project with real-time RAG and LLMOps tools, while Andrew Ng's courses are more theoretical and broad; choose this if you want hands-on deployment experience, but note it's archived and may be outdated.

Question 3

What does it cost to run the financial advisor project end-to-end?

Accepted Answer

Costs depend on cloud usage: Beam offers ~10 free GPU hours, but beyond that, you pay per use; AWS EC2 for streaming pipelines is cheap (~$0.023/hour), but external services like Alpaca and Qdrant have freemium plans that might limit scale.

Question 4

Can I swap out Falcon 7B for a different LLM like Llama 2 in this course?

Accepted Answer

Yes, the training pipeline is designed to be generic, but you'll need to adjust code for model-specific configurations and ensure compatibility with QLoRA and the tools used; the course provides a framework but may require additional customization.

Question 5

Is the Hands-on LLMs course still worth doing since it's archived?

Accepted Answer

It can be valuable for learning architecture and tools, but be prepared for potential deprecation issues; the creators recommend their new LLM Twin course for an updated experience, so check if that better suits your needs.

Question 6

How to deploy the inference pipeline without using Beam's serverless infrastructure?

Accepted Answer

You can deploy it on your own hardware if you meet the VRAM requirements, or adapt it for other cloud providers; the course focuses on Beam for simplicity, but the code is modular and can be modified with effort for alternative setups.

Hands-on LLMs: Train and Deploy a Real-time Financial Advisor

What is Hands-on LLMs: Train and Deploy a Real-time Financial Advisor?

Overview

Use Cases

Best For

Related Projects

Found a gem we're missing?

Not Ideal For

Pros & Cons

Pros

Cons

Frequently Asked Questions