Top all-time · #9432
Benjamin Marie
Personal blog · kaitchup.substack.comaiprogramming
Benjamin Marie writes The Kaitchup, a personal blog at kaitchup.substack.com about ai and programming. Since 2023, 8 different Hacker News readers have linked it 10 times, most often “Choosing a GGUF Model: K-Quants, IQ Variants, and Legacy Formats”. It is still writing: the last post was on 2026-10-31, and you can follow it by feed.
Too few sampled comments (2) to say how often HN recommends it.?
Why HN links it
HN commenters cite it for GGUF quantization evaluation results, such as the low relative error of certain quants against the base model, when discussing running large models locally.
Summarised by Claude from a sample of 2 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Weekly tutorials and news on adapting large language models (LLMs) to your tasks and hardware using the most recent techniques and models. The Kaitchup proposes a collection of 180+ AI notebooks regularly updated.
Most-linked posts?
- Choosing a GGUF Model: K-Quants, IQ Variants, and Legacy Formats2 people · 3 mentions · first linked 2026-01-10
- Falcon 180B: Can It Run on Your Computer?2 people · 2 mentions · first linked 2023-09-18
What Hacker News says?
According to this blog ( https://kaitchup.substack.com/p/lessons-from-gguf-evaluation... ) the UD_IQ2_M quants are quite strong (rel. error to the base is very low), so it's around 120GB of RAM needed, while the experts can be loaded into VRAM and the rest offloaded into system RAM. It's a high end consumer PC, sure, but not unaffordable. For example, I got an older rig with a RTX 6000 ADA (48GB VRAM), 128 GB RAM an…
Comment linking /p/lessons-from-gguf-evaluations-ternary
on “Qwen3.6-35B-A3B: Agentic Coding Power, Now Open to All”Submitted The Era of Agentic Benchmaxxing as a story · 4 points
HN discussionSome of these concerns are precisely why we are building Nemotron at NVIDIA. We want to make it possible for people to invent and deploy AI in all sorts of new and unforeseen ways. Nemotron is: 1. Not just open weight, but open data (to the limits of what is feasible), open recipe, open technique 2. In the future built by a coalition of organizations coming together to build great openly developed AI. Nemotron 3 Sup…
Comment linking /p/the-fastest-and-cheapest-120b-llm
on “The Closing of the Frontier”Submitted ThinkingCap-Qwen3.8-27B: Less Thinking, Similar Accuracy as a story · 3 points
HN discussionSubmitted GGUF Quantization Compared: Q4_K_M vs. IQ4_XS vs. IQ4_NL as a story · 2 points
HN discussionSubmitted Choosing a GGUF Model: K-Quants, IQ Variants, and Legacy Formats as a story · 2 points
HN discussionSubmitted Choosing a GGUF Model: K-Quants, IQ Variants, and Legacy Formats as a story · 2 points
HN discussionSubmitted Falcon 180B: Can It Run on Your Computer? as a story · 2 points
HN discussionSubmitted Agentic AI at Two Different Scales: Nanbeige4.2-3B and Laguna S2.1 as a story · 1 points
HN discussionSubmitted LLM Falcon 180B Needs 720GB RAM to Run as a story · 1 points
HN discussion
