strata update makes Qwen3.8-Flash-Next slower and dumber, v0.1.42.0 is the last good version
Lime-Lemon-7582 · reddit · 2026-11-15
A Reddit user reports that upgrading the local inference tool strata from v0.1.42.0 to the latest version badly degraded Qwen3.8-Flash-Next (q3s quant): a voxel pagoda scene that took 15-30 minutes (50-60 min at xhigh) now either fails after 1.5+ hours by exhausting the 262K context in endless thinking, or produces haphazard output lacking detail, repeatedly attempting overly grandiose designs then breaking them.
Multiple retries confirmed it's not a one-off; the new unsloth iq4xs quant was also a mess, so the author pins the blame on the strata version and recommends staying on 0.1.42.0. Setup: RTX 3080 10G, 64GB RAM, Linux terminal with deepseek harness. Comparison outputs for both versions are included in the post.
More from Models
- ChatGPT outage reported as users note the service is down — k7agar · 2026-11-15
- Gemini 4 Carbon Finishes ~16 Coding Tickets in 4.5 Hours, but Fable 5.2 Raises the Bar — every · 2026-11-15
- Bigger models hallucinate more, and 'hallucination is a feature,' says AI researcher — iruletheworldmo · 2026-11-15
- The recipe to solve any perception task: a few hundred examples and a fine-tuned VLM — giffmana · 2026-11-15
- AI agents go from 2.5% to 41.3% on real freelance benchmark in 13 months — kimmonismus · 2026-11-15
- Hugging Face staffer: viral 'download models now' fearmongering is misinformation — mervenoyann · 2026-11-15