5,946,648,928 bytes. That is the PTQ1_0 file of the ternary 27B everyone is posting…
This is a AI post classified by Jev as Open & local models (a model release), kept by the AI Radar because it carries real work, not commentary.
5,946,648,928 bytes. That is the PTQ1_0 file of the ternary 27B everyone is posting today, and it is 11.05 percent of the F16 file sitting in the same repo, 53,808,408,928. The 9x claim is 9.05x. Dense decode reads the whole file every token, so the ceiling is one division. Same file, four boxes: 273 GB/s, my box, 45.9 tok/s. 360, an RTX 3060, 60.5. 608, an Arc B70, 102.2. 936, an RTX 3090, 157.4. My box does 18 tok/s on the 27B Q4.5 I actually run. That is 15.2 GB per token and it lands exactly on this line, so the ternary file predicts 2.55x what the same machine gives me today. Run it o
Posted by 코지베어 🐻 CozyBear (1.2k followers) 12 h ago · 0 likes · 22 views · view the original post on X. Kept by the AI Radar as Open & local models.
More AI work like this
- dear rtx 3060 owners, and every 12gb card behind it. bonsai 2 27b dense went from 26 to… — @sudoingX
- Nimble is able to process images! — @madiator
- Woah...it claims to be open Jev. And it's 100% opensource. — @Saboo_Shubham_
- your rtx 3060 was running bonsai 2 at 26 tok/s this morning and now does 40 tok/s, i… — @sudoingX
- Meet DeepSeek-V4.1-Flash: a blazing fast image-text-to-text model. It takes both images… — @HuggingModels
- Meet Edge0-35B-A3B-preview: a 35B parameter MoE model built for edge inference. It runs… — @HuggingModels
- Putting Jeff-1 out into the world, this is an attempt at a Jev like model, built on qwen… — @DJLougen
- Train a 9M parameter language model from scratch in five minutes. — @tom_doerr
Every post is read and classified by Jev (TypeSafe): what it is, which market it belongs to, and whether the link is a real tool. 27.2k posts from 4.8k X accounts over the last 14 days, 1.2k tools, 19 markets. Collected every 5 minutes, fully re-ranked every hour — last update 2026-09-19 18:45 UTC. Full method.