- From: Dirk Schnelle-Walka <dirk@switch-consulting.de>
- Date: Thu, 4 Apr 2024 15:17:34 +0200
- To: public-voiceinteraction <public-voiceinteraction@w3.org>
- Message-ID: <1abf858b-ddd8-43f0-bc4b-55de9fccccaf@switch-consulting.de>
Dear all, in the last call we discussed a more visible representation of LLMs in https://w3c.github.io/voiceinteraction/voice%20interaction%20drafts/paArchitecture/paArchitecture-1-3.htm. I created issue #52 for this purpose: https://github.com/w3c/voiceinteraction/issues/52 I made some first attempts to have this compined. Please, check sections * 3. Architectures o image for basic architecture o descriptive text of this image * 3.2 Dialog Layer o 2 images to represent the different flavors of an IPA o descriptove text for these images * 3.2.5 Prompt adaptation o new section * 3.2.6 LLM o new section * 3.6 Resulting architecture o image for resulting architecture o description of this image Let me know if there is something that ought to be improved, maybe by asking the following questions * Do I easily get the message from the images? * Is there something in the images that confuses me? * Is there a better way to show it? * Does the description and image match my understanding of such systems? * Is there something in the text that I am missing? * Is there something in the text that confuses me? * Is there something in the text that should be removed? * Is there something in the text that should be expressed differently? Thank you Dirk
Received on Thursday, 4 April 2024 13:17:39 UTC