-
Notifications
You must be signed in to change notification settings - Fork 0
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
Showing
1 changed file
with
8 additions
and
4 deletions.
There are no files selected for viewing
12 changes: 8 additions & 4 deletions
12
content/talks/2024-09-18-amta-2024-controlling-invariants-rag.md
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -1,17 +1,21 @@ | ||
--- | ||
title: "[UPCOMING] AMTA 2024 Virtual Tutorial Day: Controlling LLM Translations of Invariant Elements with RAG" | ||
title: "AMTA 2024 Virtual Tutorial Day: Controlling LLM Translations of Invariant Elements with RAG" | ||
date: 2024-09-18 | ||
author: "ZanSara" | ||
externalLink: https://amtaweb.org/virtual-tutorial-day-program/ | ||
--- | ||
|
||
[Announcement](https://amtaweb.org/virtual-tutorial-day-program/), | ||
[slides](#), | ||
[notebook](https://colab.research.google.com/drive/1VMgK3DcVny_zTtAG_V3QSSdfSFBWAgmb?usp=sharing) and | ||
[glossary](https://docs.google.com/spreadsheets/d/1A1zk-u-RTSqBfE8LksZxihnp7KxWO7YK/edit?usp=sharing&ouid=102297935451395786183&rtpof=true&sd=true). | ||
All resources can also be found in | ||
[my archive](https://drive.google.com/drive/folders/1_8vO8O5wcvqYyjDkt2NGbwF5X6aSWgV1?usp=sharing). | ||
|
||
--- | ||
|
||
Additional backup slides on LLMs fundamentals [here](https://drive.google.com/file/d/1SN2SsMQTlA55IA8menZ4Zfkvlk8bFIzS/view?usp=sharing). | ||
_Recording coming soon._ | ||
|
||
--- | ||
|
||
At the [AMTA 2024 Virtual Tutorial Day](https://amtaweb.org/virtual-tutorial-day-program/) I talked about controlling invariant translation elements with RAG. During the talk several speakers intervened on the topic, each bringing a different perspective of it. | ||
|
||
[Georg Kirchner](https://www.linkedin.com/in/georgkirchner/) introduced the concept of invariant translation elements, such as brand names, UI elements, and corporate slogans. [Christian Lang](https://www.linkedin.com/in/christian-lang-8942b0145/) gave a comprehensive overview of the challenges of handling invariant translation elements with existing tools and how LLMs can help at various stages of the translation, covering several approaches, including RAG. Building on his overview, I showed how to implement a simple RAG system to handle these invariants properly using [Haystack](https://haystack.deepset.ai/?utm_campaign=amta-2024): we run a [Colab notebook](https://colab.research.google.com/drive/1VMgK3DcVny_zTtAG_V3QSSdfSFBWAgmb?usp=sharing) live and checked how the translation changes by introducing context about the invariants to the LLM making the translation. Last, [Bruno Bitter](https://www.linkedin.com/in/brunobitter/) gave an overview of how you can use [Blackbird](https://www.blackbird.io/) to integrate a system like this with existing CAT tools and manage the whole lifecycle of content translation. |