News for hackers. Stuff that matters.

A comment on Qwen 3.8-Flash-Next releasing tomorrow (125B a6B). Parent comment

Re:Qwen 3.8-Flash-Next releasing tomorrow (125B a6B) by SwellJoe on Tuesday August 25, 2026 @05:33PM (#49440969)
What model? Also, I don't know what "the PrismaAQUA" means, ddg thinks it's a CPAP machine, which seems unlikely to help with inference performance.

Also, 4-bit has measurable intelligence loss. Sometimes worth it, but, at this size models are barely smart enough at 8 or 6.

[ Parent | on HN ]
Re:Qwen 3.8-Flash-Next releasing tomorrow (125B a6B) by decide1000 on Tuesday August 25, 2026 @07:04PM (#49441888)
Qwen3.8-27B-PrismaAQUA-5.5bit-vllm

The output quality is higher. It's held at full precision (not quantized).

[ Parent | on HN ]