Someone squeezed a Qwen3.8-27B fine-tune down to ~7.4GiB on the GPU. Not sure I believe…
This is a dev post classified by AGK Intelligence as AI dev tools (a tool drop), kept by the Dev Radar because it carries real work, not commentary.
Someone squeezed a Qwen3.8-27B fine-tune down to ~7.4GiB on the GPU. Not sure I believe it. And somehow it just scored higher than Qwen3.6-27B BF16 on LiveCodeBench. 👀 The new Swift 1.5 / Qwen3.8-27B ATX AI2 build is only ~2.3 bits per weight. Stats 💾 ~7.4GiB resident on GPU incl. MTP 🧠 27B parameters ⚡ >90 t/s reported for agent work on 3090/3080 Ti 📚 designed to make 200K+ context possible on 12GB Ampere cards ⚙️ custom llamAmpere runtime This is almost unbelievable. LiveCodeBench v6: Qwen3.8 27B 2.3bpw → 84.3% Qwen3.6 27B BF16 → 83.9% Qwen3.8 27B BF16 → 90.3% So no, the 2.3-bit model
Posted by David Hendrickson (12.6k followers) 1 h ago · 6 likes · 469 views · view the original post on X. Kept by the Dev Radar as AI dev tools.
More dev work like this
- “The Matrix is everywhere. It is all around us.” — Morpheus — @mishig25
- Meet Apriel-Voice-27B-preview: a new audio-text-to-text model from ServiceNow-AI. It… — @HuggingModels
- Stop rebuilding memory and tool wiring for every agent. — @DanKornas
- 😱卧槽！万物皆可逆向！Rust神器craft-anything炸场，看懂任何闭源软件，AI直接帮你重造一个自己的版本！ — @NFT_Chen
- I look at the GBrain Tools screen quite a bit to see what all my agents are up to during… — @bradgessler
- ComfyUI 玩家这回能直接抄作业了，有人把 MiniMax H3 的“导演台”做好了。 — @bkdgiffug
- Just hit 238 followers — thank you 🙏 — @mims3590
- Introducing SubQ Snap 1.1 on Baseten. — @baseten
Every post is read and classified by AGK Intelligence: what it is, which market it belongs to, and whether the link is a real tool. 31.3k posts from 5k X accounts over the last 21 days, 9.1k tools, 12 markets. New posts are collected several times an hour and the full ranking is recomputed every hour — last update 2026-11-02 21:22 UTC. Full method.