Best ToGo G๐˜‚๐—ถ๐—ฑ๐—ฒ ๐—ณ๐—ผ๐—ฟ ๐—š๐—ฒ๐—ป๐—ฒ๐—ฟ๐—ฎ๐˜๐—ถ๐˜ƒ๐—ฒ ๐—”๐—œ ๐Ÿง 

Amber Ivanna Trujillo
2 min readApr 27, 2024

Pondering whether to jump into prompt tinkering, RAG adventures, fine-tuning galore, or birthing your very own model from scratch?

Feast your eyes on one of my absolute faves, the diagram of dreams that I whip out to dazzle clients, where I spend most of my time strutting my stuff.

Iโ€™ve dished out some juicy tips to help you navigate the maze of options, like a pro gamer leveling up their character.

The whole process stays pretty much the same, but knowing your way around these choices is the real key to unlocking the treasure chest.

With LLMs getting better by the minute and new models popping up like daisies, itโ€™s super important to pick the right path. Do you go for the top dog all-rounder or the slightly less flashy but super specialized model tailored just for you?

Hereโ€™s the scoop on how to do it:

  1. Zero Chill: Take the model as it comes, no fuss.
  2. Prompt Party: Start simple and ramp up the complexity as you go.
  3. Basic RAGtime: Perfect for when you need quick answers and crystal-clear explanations, made easy peasy with โ€œKnowledge Basesโ€ in Amazon Bedrock.
  4. Advanced RAGtime: For when youโ€™re swimming in data, youโ€™ll need the big guns.

--

--

Amber Ivanna Trujillo
Amber Ivanna Trujillo

Written by Amber Ivanna Trujillo

I am Executive Data Science Manager. Interested in Deep Learning, LLM, Startup, AI-Influencer, Technical stuff, Interviews and much more!!!

Responses (2)