Moefit: Fit Qwen 3.8 Flash on your Mac with limited ram (github.com via hn) moeqwen 5w
event
LLM Offloading Event
-
moefit Run Qwen3.8-Flash-Next on a Mac that does not have enough RAM for the full model. Point your coding agent at this repo, give it the setup prompt below, and it can install and run a ~125B MoE model on Apple Silicon with 24–64 GB of u…
-
I made an "Opus 5.5 style" code-rendered video — but on a local NVFP4 Qwen 3.8 27B (www.reddit.comhttps) qwenopus 5w
The "Opus 5.5 makes videos" thing has been going around — and the interesting part is that it isn't generating pixels. The model writes a self-contained HTML/Canvas scene where every frame is a pure function of time, a headless browser cap…
-
Native memory for local LLMs, (www.reddit.com via reddit) vllmqwen 5w
TL;DR: Native consumption of memory at the LLM level, no context. It's generally applicable to transformer-based models as well as Mamba and similar architectures.
-
at q4 can i run qwen 3.8 flash next with strata on these specs? specs in desc (www.reddit.com via reddit) qwen 5w
24gb gddr7 2287 MHz 96gb ddr5 6400mhz core ultra 9 285hx, 5.5 ghz
-
Qwen3.8 Flash Next on 5060 Ti 16GB - 55 tok/s average, and a few demos (www.reddit.com via reddit) 5w
Hello guys! I've been out of the loop for a while.
-
-
Show HN: Ototo: Offload code exploration to a smaller model from Claude et al. (ototo.dev via hn) qwen 5w
Hey all - I made a thing with my clanker Kevin: Ototo (https://ototo.dev/) - yes this is clanker driven dev. The idea is to offload code exploration and other token'y heavy tasks to a smaller model to do the work and chuck the result back.
-
Jeff-Code: a 0.8B model makes Qwen 3.8-27B coding 47% faster, same pass rate (github.com via hn) qwen 5w
Jeff-Code A coding agent where a 0.8B decision model works alongside Qwen 3.8-27B. Coding tasks 47% faster (32% less time) on average, at the same pass rate.
-
The Rise of Overfit Inference Engines (carteakey.dev via hn) moe 6w
The surprise For a week I'd been tuning Qwen3.8-Flash-Next, a 125B-parameter MoE, on 🧠 yeti-cachy (my main homelab node: an i5-12600K with 64 GB DDR5 and an RTX 4070 12GB I paid $500 for). After a lot of flags and two experimental branches…
-
I made my iPhone a second GPU for my MacBook-Qwen 3.8 27B prefills 29–44% faster (reddit.com via hn) qwen 6w
could not extract summary
-
Faster Local LLMs with iPhone Offloading (github.com via hn) 6w
Backburner Plug your iPhone into your MacBook with a 10 Gb/s USB-C cable and it helps run Qwen3.8-27B locally: Faster prefill (up to 64k context). For every batch of prompt tokens the Mac runs layers 1-40 and the iPhone runs 41-64 on its G…