FLM-Audio: Natural Monologues Improves Native Full-Duplex Chatbots via Dual Training · HackerPapers discussionshttps://hackerpaper.kaipfister.de/p/2509.02521/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:27:26 GMTFLM-Audio: Natural Monologues Improves Native Full-Duplex Chatbots via Dual Training — HN submissionhttps://news.ycombinator.com/item?id=45299218urn:hackerpapers:submission:2509.02521:45299218Fri, 19 Sep 2025 08:14:44 GMT<p>Show HN: FLM- Audio, our 7B spoken dialog chatbot with native full-duplexity</p><p>We’re excited to share RoboBrain-Audio（FLM-Audio）, a new 7B spoken dialog chatbot with native full-duplexity. FLM-Audio achieves superior response qualities and chatting experiences while requiring significantly less training data. Key innovations: Natural Monologue Abandons word-level timestamps and innovatively proposes the “Natural Monologue ” mechanism Preserves the inherent advantages of LLMs in generating coherence and instruction-following. effectively addresses the context-dependent pronunciation issues of certain words (especially numbers). Dual Training Paradigm Training spans two…</p><p>2 points; 0 discussion replies on HN. <a href="https://hackerpaper.kaipfister.de/p/2509.02521/">Paper and source history</a>. A submission is not an endorsement.</p>cs.SD