Statistically Lossless Quantized GGUF of Qwen3.8 27B Released
kaitchup released a statistically lossless quantization (SLQ) GGUF version of Qwen3.8 27B on Hugging Face, about 23GB and reportedly close to FP8 quality with 99.5% fidelity for UD Q8. Reddit users tested it and discussed the hidden paywall in the repo.
2026-10-27 ~ 2026-10-28 · 3 related posts
- Statistically lossless Qwen3.8 27B GGUF quants drop, UD Q8 matches BF16 at 99.5% — udmrzn · 2026-10-27
- Statistically Lossless Qwen3.8 27B GGUF Quants Hit ~99% EAR at 23GB, Feels FP8-like — Dany0 · 2026-10-28
- Statistically lossless Qwen3.8-27B GGUF quantizations land on Hugging Face, analysis paywalled — Dany0 · 2026-10-28