Deep Learning and LLM

Equip students to evaluate, fine‑tune, and productionise LLMs via transfer learning and Hugging Face.

Get Course Info

Audience: Developers / Data Scientists / Team Leads / PMs

Duration: 2 days

Format: Lectures and hands‑on labs (50% lecture, 50% lab)

Overview

Large Language Models (LLM) are transforming industry. This course focuses on open‑source models, Hugging Face libraries, and practical techniques for evaluating, fine‑tuning, and deploying LLMs in production environments.

Objective

Equip students to evaluate, fine‑tune, and productionise LLMs via transfer learning and Hugging Face.

What You Will Learn

Evaluate LLM models & claims
Hugging Face for model hosting & deployment
Prompt engineering best practices
Transfer learning, pre‑training, fine‑tuning
Packaging models for deployment
Creating & improving ML pipelines in production

Course Details

Audience: Developers / Data Scientists / Team Leads / PMs

Duration: 2 days

Format: Lectures and hands‑on labs (50% lecture, 50% lab)

Prerequisites:

General familiarity with machine learning

Setup: Zero‑install cloud lab • Modern laptop • Chrome browser

Detailed Outline

DL use‑cases, neural‑network families (NN, CNN, RNN)

Best practices & practical advice

Transformers library, models, pipelines
Putting it all together

Data processing
Trainer API / Keras workflows
Full training cycle

Token classification, masked LM, translation, summarisation, causal LM
Question answering

Overview & comparison of open LLMs
Evaluating & fine‑tuning LLMs
Alpaca, LLaMA, Dolly, Nomic, Vicuna