Guide to using the language power of Large Language Models for your work on the AX NLG
Writing with Large Language Models such as GPT, ChatGPT or BERT seems to be intuitive and simple: You create a prompt by entering keywords, whole sentences or a text briefing, or you provide the LLM with a text example and instruct it to paraphrase or rewrite an issue. Until now, it was always a bit of trial and error before you finally received a text that was suitable for your needs. Now there are sophisticated strategies for composing the writing task.
Integrating LLMs into data-to-text work using the AX NLG platform goes one step further. Having the entire text created with the prompt optimization strategies on ChatGPT, for example, and then integrating it into your text project can be very time-consuming. A better approach is to use the linguistic and creative capabilities of the platform at appropriate points in the AX project process.
What you will learn in this guide
- You have an overview of how language models work.
- You know at which point in the writing process you can use them.
- You are able to work with the integrated LLMs of the NLG platform
Prerequisites: This is the world of LLMs and writing
How do LLMs work?
What are their strengths, and what are their weaknesses?
What are the benefits of using LLMs for Copy Writing?
What are the drawbacks?
Adjusting your expectations to get the best out of LLMs
Data-to-text and LLMs different approaches to Scaling
The best of two worlds: Working with LLMs on the NLG Platformm
Statement Suggestions – helps you to keep the story going
- Based on your existing text along with your variables, the GPT-powered writing tool suggest the perfect follow-up statement
- You can get statement suggestions from scratch, based on your product data
Alternative Text Suggestions
With a single click, you can explore up to 10 creative, GPT-generated alternatives for any selected text using the intuitive inline toolbar. In the next step distinct branches are generated for each alternative.
- Get up to 10 alternatives for any selected text
- Create branches for each alternative at the touch of a button and choose the perfect fit for your content
Automated Translation with LLMs
Data Privacy
The GPT API allows us to generate text based on your text, data, containers, and variables. This ensures that you only get customized text that matches your content. However, if you wish to exclude your project data from external services such as OpenAI, you can easily opt out.
How to get the ML-Powered Suggestion Tools?
EnableML-powered Suggestions
in your Advanced Settings. If you don't want to use the new GPT features, you can easily opt out by unchecking the External Service Providers boxes.
Based on GPT
Both our new ML-powered features are backed by OpenAI's GPT API, one of the most advanced large language models available today. The GPT API lets us generate text based on your project's active test object, variables and the text you have already written. This means that the suggestions you get are tailored to your project and the text you have already written. In contrast to ChatGPT, OpenAI will not use API data to train OpenAI models (see their API data usage policies). OpenAI may receive rendered content of statements, name and output value of variables, and contents of the selected test object. They will not receive any intermediate values and ruleset code. Should you still not want us to send your project data to OpenAI, you can opt-out of OpenAI or deepL features per project in your project settings.
TIP
Container or variable suggestions are still active when this services are disabled.
Available Languages
ML-powered suggestions are currently available for the following languages:
Catalan (ca-ES), Chinese (zh-), Croatian (hr-), Danish (da-DK), Dutch (nl-), English (en-), Finnish (fi-FI), French (fr-), German (de-), Greek (el-GR), Italian (it-IT), Japanese (ja-JP), Korean (ko-KO), Latin (la-VA), Lithuanian (lt-LT), Macedonian (mk-MK), Norwegian Bokmål (nb-NO), Polish (pl-PL), Portuguese (pt-), Romanian (ro-RO), Russian (ru-RU), Serbian (sr-RS), Slovenian (sl-SI), Spanish (es-), Swedish (sv-SE), Ukrainian (uk-UA).
What are GPT-3 and Data-to-Text and how do they differ?
What do you want to do next?
Video: Saim Alkan explains https://youtu.be/aTlP_LkVOrQ