{"slug": "ibm-research-trains-20b-model-on-its-design-language-for-slide-decks", "title": "IBM Research trains 20B model on its design language for slide decks", "summary": "IBM Research researchers Praveen Venkateswaran, Anupama Murthi and Hamid Adebayo published Palette on November 3rd, 2026, a 20B open-weight model fine-tuned on synthetic slide decks built around the IBM Design Language that plans, codes and repairs each slide. The approach accepts a repair only when a geometry checker confirms the rendered slide improved, and IBM positions the self-hosted model as a way for companies to keep confidential material inside their own systems while producing decks that match their own design language.", "body_md": "# IBM Research trains 20B model on its design language for slide decks\n\n**Praveen Venkateswaran, Anupama Murthi and Hamid Adebayo fine-tuned a 20B model on synthetic decks following IBM's visual system, then used a geometry checker to keep a repair only when the rendered slide improved.**\n\n        By [RuntimeWire Staff](https://runtimewire.com/author/runtimewire-staff)\n        · Published \n\nPrimary source: [Hugging Face Newsroom](https://huggingface.co/blog/ibm-research/palette-deck-building)\n\n## Why it matters\n\nPalette treats enterprise design consistency as a training and verification problem. A self-hosted 20B model that plans, codes and repairs each slide could let companies keep confidential material inside their own systems while still getting decks that look like their own, provided the design-language claim holds beyond IBM's template.\n\nPraveen Venkateswaran, Anupama Murthi and Hamid Adebayo of IBM Research published Palette on [November 3rd, 2026](https://huggingface.co/blog/ibm-research/palette-deck-building?ref=runtimewire), describing a 20B open-weight model fine-tuned on synthetic slide decks built around the [IBM Design Language](https://www.ibm.com/design/language/resources/?ref=runtimewire). The practical claim is narrow and testable: train a smaller model on one visual system, give it an intermediate plan to edit, and check each rendered slide before accepting a repair.…\n\nWhen you purchase through links in our articles, [we may earn a small commission](https://runtimewire.com/affiliate-monetization-standards). This doesn’t affect our editorial independence.", "url": "https://wpnews.pro/news/ibm-research-trains-20b-model-on-its-design-language-for-slide-decks", "canonical_source": "https://runtimewire.com/article/ibm-research-palette-20b-deck-model-design-language", "published_at": "2026-11-03 16:44:16+00:00", "updated_at": "2026-11-03 19:49:20.790883+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-research", "generative-ai", "ai-tools"], "entities": ["IBM Research", "Palette", "Praveen Venkateswaran", "Anupama Murthi", "Hamid Adebayo", "IBM Design Language", "Hugging Face"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/ibm-research-trains-20b-model-on-its-design-language-for-slide-decks", "markdown": "https://wpnews.pro/news/ibm-research-trains-20b-model-on-its-design-language-for-slide-decks.md", "text": "https://wpnews.pro/news/ibm-research-trains-20b-model-on-its-design-language-for-slide-decks.txt", "jsonld": "https://wpnews.pro/news/ibm-research-trains-20b-model-on-its-design-language-for-slide-decks.jsonld"}}