Large Language Models for Real-World Applications - A Gentle Intro
summary
Clear, concrete rundown: shows how GPT3 and Dall‑E class models are trained and deployed, demonstrates embeddings and fine‑tuning for classifiers and semantic search, and runs image VQA demos (Flamingo) plus image generation (Stable Diffusion). Covers prompt tactics—temperature, multiple generations, scratchpad/chain‑of‑thought—and practical deployment notes (Cohere API, Triton, reduced precision, distillation/adapters). Claims work at product scale (search snippets, chatbot intent, content moderation) and gives actionable pitfalls and data‑augmentation recipes.
AI-written summary · about
chapters (10)
- 0:00Welcome to LLMs
- 3:56Real-world applications
- 8:28Visual understanding with LLMs
- 13:19Why LLMs matter
- 20:44Understanding embeddings
- 27:04Building classifiers
- 33:06Prompt engineering
- 42:10Text pipelines & Q&A
- 44:49Summary & demo
- 46:35Q&A
related talks
36mLarge Language Models for Real-World Applications - A Gentle Intro
43mBeyond the Hype: A Realistic Look at Large Language Models
30mLarge Language Models and Prompt Engineering for Developers
01h 00mInside GPT – Large Language Models Demystified
47mInside GPT – Large Language Models Demystified
01h 00mUnleashGPT: Mastering the Implementation of Large Language Models
27mEmbeddings, Transfer Learning and More: LLM Tailored to You
56mLarge Language Models for Python Specialists