Prompt Engineering and LLMOps: Tips and Tricks
8Here, I share some tips and tricks I have for a new emerging field of LLMOps - Interfacing LLMs in a systems-level approach. This has powered emerging frameworks such as AutoGPT and BabyAGI. I also share some tips I have for prompt engineering, including zero-shot, few-shot and conditional generation on earlier prompts.
~~~~~~~~~~~~~~~~~~~~~~~~~~~
References:
Slides: https://github.com/tanchongmin/TensorFlow-Implementations/blob/main/Paper_Reviews/Prompt%20Engineering.pdf
Generative Agents: https://arxiv.org/abs/2304.03442
Paper which Ze Yu mentioned about LLMs (The False Promise of Imitating Proprietary LLMs): https://arxiv.org/pdf/2305.15717.pdf
BabyAGI: https://github.com/yoheinakajima/babyagi
Gandalf Game: https://gandalf.lakera.ai/
Zero-shot prompting: https://www.youtube.com/watch?v=C0Eug9XpcBo
ARC Challenge: https://www.youtube.com/watch?v=vt2yG1da8Fg
~~~~~~~~~~~~~~~~~~~~~~~~~~~
0:00 Introduction
1:07 LLMs vs Computer Programs
15:13 Key benefits of LLMs
16:18 Systems-level approach to LLMs
19:32 BabyAGI
24:50 Gandalf Game
28:05 Zero-shot prompting
35:00 Few-shot prompting
43:12 How to program intent?
46:25 Consistent Output
55:20 Asking LLM to fill up outputs
56:07 Memory to augment context (as API, or by repeating regularly via prompt)
57:41 Sequential Prompting
1:00:58 Breakdown of Prompt for Court Simulator Game
~~~~~~~~~~~~~~~~~~~~~~~~~~~
AI and ML enthusiast. Likes to think about the essences behind breakthroughs of AI and explain it in a simple and relatable way. Also, I am an avid game creator.
Discord: https://discord.gg/bzp87AHJy5
LinkedIn: https://www.linkedin.com/in/chong-min-tan-94652288/
Online AI blog: https://delvingintotech.wordpress.com/
Twitter: https://twitter.com/johntanchongmin
Try out my games here: https://simmer.io/@chongmin