Feed
Pro
Search

Author

Write
Drafts

Bug0 - The AI-native e2e QA regression testing Passmark - The open-source AI framework for regression testing Hackathons Changelog Brand Hashnode gql skill - let your AI agent publish to your Hashnode blog The Foreword by Hashnode - official blog from the Hashnode team @hashnode on X Hashnode on LinkedIn Support - hello+support@hashnode.com Code of Conduct Terms Privacy Sitemap
Sign in

Search Hashnode

Search posts, tags, users, and pages

@joshgreen

Josh Green

@joshgreen·Budapest, Hungary·Joined June 2026

Lead developer at ModelDirectory.org

About

I frequently automate problems that should not be automated. I then often brag about it.

Available for

Freelance frontend work, technical writing, open source collaboration

Josh Green's blogs

Bench Notesjoshgreen.hashnode.dev3 posts

About

I frequently automate problems that should not be automated. I then often brag about it.

Available for

Freelance frontend work, technical writing, open source collaboration

Josh Green's blogs

Bench Notesjoshgreen.hashnode.dev3 posts

Articles Comments

Recently published

JGJosh Greeninjoshgreen.hashnode.dev·3d ago · 7 min read

Your Local MoE Model Is Slow: A Diagnostic Checklist

Mixture-of-experts models are the reason a 30B can decode like a 3B on a desktop. They are also the reason people post "why is my local model crawling" threads, because MoE has a handful of failure mo

JGJosh Greeninjoshgreen.hashnode.dev·6d ago · 8 min read

GLM 4.7 Flash on Two RTX 3090s: The Harness, Every Config, and a Decision Table

I keep two RTX 3090s in a desktop, not a server. No NVLink, a consumer motherboard, the kind of build a lot of people actually have sitting under a desk. The question I wanted answered was narrow and

JGJosh Greeninjoshgreen.hashnode.dev·Jun 6 · 8 min read

Why DDR5 Bandwidth Kills Dual-LLM Inference on APUs (Benchmarks Inside)

Did you know that a 35-billion-parameter model can generate tokens at the same compute cost as a 4B model? That single fact made me abandon a multi-model agent architecture I'd spent a weekend buildin