rguiscard

Joined 7 years ago

Karma

3,885

Submissions

467

Member since

Mar 6

[ my public key: https://keybase.io/rguiscard; my proof: https://keybase.io/rguiscard/sigs/mVHEOkVvrO76COiMnXbWWNeb7m_dIYLKst5PcEKWXcM ]

View on Hacker News

Recent submissions

碁の国: Play Chess as Town Building gonokuni.plpl.town

12 rguiscard 3 days ago 3 comments

FreeBSD 16.0-CURRENT/aarch64 guest boots on Apple Virtualization.framework github.com

5 rguiscard 9 days ago 1 comments

EU negotiates sanction swap with China to rescue critical ASML supplier tomshardware.com

3 rguiscard 13 days ago 0 comments

China startup demoes prototype EUV light source, still years away from viability tomshardware.com

2 rguiscard 13 days ago 0 comments

BeBox is a Haiku application that emulates Be Inc.'s BeBox github.com

94 rguiscard 15 days ago 34 comments

RenkuOS: Haiku OS forked with AI-assisted coding and LLM integrations renkuos.org

2 rguiscard 16 days ago 1 comments

Prism-ML Bonsai 2 Joins Our (ByteShape) Qwen3.8 Quantization Comparison reddit.com

1 rguiscard 23 days ago 0 comments

Longest traffic jam (number of vehicles): Berlin 1990 guinnessworldrecords.com

3 rguiscard 24 days ago 0 comments

Huawei Kirin 9050 Pro uses impressive 3D V-Cache-like stacked design for the SoC notebookcheck.net

1 rguiscard 26 days ago 0 comments

Free Your Bios Today gnu.org

3 rguiscard 26 days ago 0 comments

Vfkit: CLI tool to start virtual machines using macOS Virtualization framework github.com

1 rguiscard 1 month ago 0 comments

ECB will invest in tokenised securities, and settle trades on its own new rail thenextweb.com

14 rguiscard 1 month ago 0 comments

Bsdkrun: A Firecracker-style MicroVM for BSD/Linux guests, built on libkrun github.com

1 rguiscard 1 month ago 0 comments

Radar: An Expert-Level Generalist AI for Abdominal CT Diagnosis github.com

3 rguiscard 1 month ago 0 comments

Comment · 1 month ago

I wonder the same thing for Bonsai 2. ByteShape offers 5 models from IQ2_XXS-2.56bpw (8.8GB), IQ3_XXS-2.88bpw (9.9GB), IQ3_XS-3.01bpw (10.4GB), IQ3_S-3.23bpw (11.0GB) to IQ4_XS-3.84bpw (13.1GB). Their benchmarks show gradual improvement with size and users can pick one to fit theirs need. Bonsai-2-27B now is about 8.6GB. It might be good to have a quaternary version around 10-11GB to fit a computer with 16-24GB RAM.