The quantized builds run on machines with 64GB+ of RAM.
Unsloth has released new GGUF builds of DeepSeek-V4.1-Flash-1017 for running on local machines.
Unsloth says its Dynamic V3.1 quantizations lead across nearly all benchmarks it tested, including GPQA Diamond, AIME25, LiveCodeBench v6, MMLU-Pro and Aider.
The model runs on 64GB or more of RAM. Download links and a blog post are available.
Each commit is a real law taking effect, so you can see what the law said on any date.
@RoundtableSpace1.7k
Xiaomi-OCR-0 handles parsing, extraction and OCR questions in one small model.
@XiaomiTech_1.6k
SenseTime says the 9B model performs like a 27B model on hard tasks.
@SenseTime_AI647
A free, open-source tool turns a machine into a self-managing simulator cloud.
@ChmalSzymon632
Blindsight is meant to let blind people, including people who have never seen, see.
@elonmusk67k
Musk says it lets people control computers just by thinking.
@elonmusk43kOne email each morning with the AI, crypto and tech posts everyone is talking about, explained in plain English.
