Introduction¶
About this Guide
This guide is in an initial early access state currently and is written to organize my play time as I play with, and apply, these tools.
The content is about solving real problems e.g. how to
- view the main topics in a set of documents
- validate assigned CWEs, and suggest CWEs to assign
- chat with large documents
- extract configuration parameters from user manuals.
These examples were driven by a user need.
While the examples focus on specific areas, they can be applied in general to many areas.
After reading this guide you should be able to
- Apply Language Models to augment and amplify your skills.
- Understand the types of problems that suit Language Models, and those that don't
Overview¶
Intended Audience¶
The intended audience is people wanting to go beyond the hype and basics of Large Language Models.
No prior knowledge is assumed to read the guide - it provides just enough information to understand the advanced topics covered.
A basic knowledge of Jupyter Python is required to run the code (with the data provided or on your data).
How to Use This Guide¶
How to Contribute to This Guide¶
You can contribute content or suggest changes:
- 💬 Discuss content
- 💡 Suggest content
- 🪲 Report Errors, typos
Writing Style¶
The "writing style" in this guide is succinct, and leads with an opinion, with data and code to back it up i.e. data analysis plots (with source code where possible) and observations and takeaways that you can assess - and apply to your data and environment. This allows the reader to assess the opinion and the code/data and rationale behind it.
Different, and especially opposite, opinions with the data to back them up, are especially welcome! - and will help shape this guide.
Quote
If we have data, let’s look at data. If all we have are opinions, let’s go with mine.
Notes¶
Notes
- This guide is not affiliated with any Tool/Company/Vendor/Standard/Forum/Data source.
- Mention of a vendor in this guide is not a recommendation or endorsement of that vendor.
- This guide is a living document i.e. it will change and grow over time - with your input.
This guide is not about which tool is better than the other
"Don't fall in love with models: they're expendable. Fall in love with data!"
Julien Simon, Chief Evangelist, Hugging Face
Warning
You are responsible for your data and where it goes.
If you don't understand where your data goes, and what happens to it for a given model or tool, then find out before you use private or personal data.
To evaluate models and tools, you can start with public data.