Unsloth Releases Qwen3.8-27B GGUFs with 10% Higher Accuracy
danielhanchen · reddit · 2026-08-20
Unsloth released new GGUFs for Qwen3.8-27B using Dynamic v3.0, offering 10% higher accuracy at the same size and outperforming others by >10% on Div-300 and KLD benchmarks. They also released 1-bit quants retaining 77% accuracy, runnable on 8GB RAM. The method relies solely on post-training quantization without QAT/QAD or imatrix dataset training. An overfitting analysis and the imatrix file are provided for the community. The Unsloth Desktop will also update with auto-compaction and external API tool calling.
Related event: Unsloth's Dynamic v3.0 Quantization Brings Qwen3.8-27B to 8GB Machines(5 posts)→
More from Infra
- Large data centers employ only ~50-150 permanent on-site staff, Microsoft says — brezina · 2026-10-16
- Prediction: Opus 5.5-level models on laptops by 2027, backed by Qwen 27B data — cephaloform · 2026-10-16
- Oracle down 12% in a week, nearly 50% off June high as OpenAI exposure spooks market — alvelda · 2026-10-16
- US data center moratoriums top 400 — an AI bull explains why they're actually good for AI — kevinsxu · 2026-10-16
- Microsoft Research unveils Magnetar, the fastest zkVM at 10-39x speedups and post-quantum security — jedisct1 · 2026-10-16
- Bonsai 2 2B hits 120 tok/s on iPhone 18 Pro, outpaces Qwen 3.5 2B at half the memory — cephaloform · 2026-10-16