DFM Mimir v1: Open 1B Model Achieves SOTA Danish Performance
SDU-Denmark · hf · 2026-08-17
SDU Denmark releases Mimir v1, a 1B-parameter hierarchical reasoning model trained on permissible data, achieving competitive English and SOTA Danish results.
More from Models
- Users Report ChatGPT Getting Noticeably Dumber at Following Instructions in Recent Weeks — Cubic_Inequality504 · 2026-08-17
- DeepSeek V4 Flash tested on Mac Studio: Impressive quality, high RAM demand — pj-frey · 2026-08-17
- GPTZero flags pure ChatGPT text as "Mixed" — No-Mortgage-2037 · 2026-08-17
- User Review: Gemini 3.7 Flash Shows Noticeable Improvement Over 3.6 — gabriberton · 2026-08-17
- Heavy user complains: Opus 5 and Fable 5 experience has regressed significantly — madscene · 2026-08-17
- Local Object Detection with Qwen3.8-27B: Cross-Validating with RF-DETR — MaziyarPanahi · 2026-08-17