Qwen 27B NVFP4 Quantization Underperforms Baseline
Pyrolistical · reddit · 2026-09-01
The author evaluated Qwen3.8 27B NVFP4 GGUF models against Unsloth baselines using perplexity scores. Results show that all tested NVFP4 models have worse perplexity for their file size compared to standard quants, even underperforming Q4KXL. This likely explains why Unsloth does not publish NVFP4 GGUF versions.
More from Models
- MOSS SFT vs Flux 3 Klein vs SD 3.0: Same-Prompt Comparison Exposes SD3's Age — ZootAllures9111 · 2026-10-18
- Quick take: GPT-6 matches Opus 5.5 at coding, far ahead at computer use — _vztu · 2026-10-18
- Open models now above Sol-level, outpacing expectations as community awaits V4.1 Pro — yacineMTB · 2026-10-18
- Codex took 3 days, Opus 5.5 nailed it in an afternoon — banned user's video protest — dotey · 2026-10-18
- Yacine MTB: Users will remember how Anthropic treated them once intelligence proliferates — yacineMTB · 2026-10-18
- Mistral's Timestral forecasting models start at 5M params and run on plain CPUs — lmoroney · 2026-10-18