SomeOddCodeGuy

Just a boring dev/manager who picked up LLMs in early 2023, and been engrossed ever since. This is a cross-post blog. The main blog is at https://someoddcodeguy.dev

Florida, USA Joined on Sep 9, 2025 https://github.com/someoddcodeguy

SomeOddCodeGuy

Oct 13

Understanding MoE Offloading

#deeplearning #llm #architecture #performance

5 min read

SomeOddCodeGuy

Oct 10

My Next Steps with Wilmer

#architecture #devjournal #llm

4 min read

SomeOddCodeGuy

Oct 10

Microsoft's New User Role Model

#microsoft #testing #llm #tooling

1 min read

SomeOddCodeGuy

Oct 5

Agentic Coding Pt 2...

2 min read

SomeOddCodeGuy

Oct 5

Reddit woes; but a light at the end of the tunnel

1 min read

SomeOddCodeGuy

Oct 1

GLM 4.6 MXFP4 vs q8_0 gguf speeds on Mac M3 Ultra

2 min read

SomeOddCodeGuy

Sep 28

I'm Not Sure If This Counts As "Vibe Coding"?

2 min read

SomeOddCodeGuy

Sep 27

Recursive Workflows- Offline Wikipedia Search in WilmerAI

3 min read

SomeOddCodeGuy

Sep 23

Stanford Lectures...

1 min read

SomeOddCodeGuy

Sep 20

Heh... Oops. All Hail the Unit Tests.

1 min read

SomeOddCodeGuy

Sep 14

My Unorthodox Homelab Setup: Updated

#showdev #devops #tooling

5 min read

SomeOddCodeGuy

Sep 11

Latency while using RDP

1 min read

SomeOddCodeGuy

Sep 10

Time For Another Studio? Wallet Beware...

1 min read

SomeOddCodeGuy

Sep 9

A Quick List of LLM Benchmarks

1 min read

SomeOddCodeGuy

Sep 4

RAG Really Is More of a Software Problem Than An AI Problem

2 min read

SomeOddCodeGuy

Aug 29

That LinkedIn '95% of AI Ventures Fail' Stat That's Going Around...

2 min read

SomeOddCodeGuy

Aug 25

Mac Studio M3 Ultra Speeds for Qwen3 235b, GPT-OSS-120b, GLM 4.5, and Deepseek V3.1

2 min read

SomeOddCodeGuy

Aug 7

Reddit Shadowbans- A Deep Dive Into What Little I Could Find

4 min read

SomeOddCodeGuy

Jun 26

I'll Always Have A Softspot for the Old Text-Generation-WebUI Chat Bubbles

1 min read

SomeOddCodeGuy

May 29

Running Deepseek R1 0528 q4_K_M and mlx 4-bit on a Mac Studio M3

2 min read

SomeOddCodeGuy

Mar 25

M3 Ultra Mac Studio 512GB prompt and write speeds for Deepseek V3 0 671b gguf q4_K_M, for those curious

2 min read

SomeOddCodeGuy

Mar 15

Running Llama 3.1 405b q6 and Command-A 111b Q8 on M3 Ultra Mac Studio

1 min read

SomeOddCodeGuy

Mar 12

Mac Speed Comparison: M2 Ultra vs M3 Ultra using KoboldCpp

3 min read

SomeOddCodeGuy

Sep 24 '24

Low Context Speed Comparison: Macbook, Mac Studios, and RTX 4090

3 min read

SomeOddCodeGuy

Sep 6 '24

My Personal Guide for Developing Software with AI Assistance: Part 2

6 min read

SomeOddCodeGuy

Jul 5 '24

MMLU-Pro Combined Results - Model Quantization Comparison

13 min read

SomeOddCodeGuy

Jun 22 '24

Meet WilmerAI- my open source project to maximize the potential of Local LLMs via prompt routing and multi-model workflows

9 min read

SomeOddCodeGuy

May 27 '24

Offline Wikipedia API- An easy to use offline API that serves up full text Wikipedia articles.

3 min read

SomeOddCodeGuy

May 18 '24

My Personal Guide for Developing Software with AI Assistance

7 min read

SomeOddCodeGuy

May 17 '24

Almost a year later, I can finally do this. A small teaser of a project I'm working on

2 min read

SomeOddCodeGuy

Mar 8 '24

Frankenmerges are actually kind of great...

2 min read

SomeOddCodeGuy

Mar 2 '24

Ok, I admit- SillyTavern is a great way to test models after all

3 min read

SomeOddCodeGuy

Feb 19 '24

Real World Speeds on the Mac: Koboldcpp Context Shift Edition!

9 min read

SomeOddCodeGuy

Feb 17 '24

Here Are Some Real World Speeds For the Mac M2 Ultra, In Case You Were Curious

6 min read

SomeOddCodeGuy

Jan 19 '24

Deepseek 67b is amazing, and in at least 1 usecase it seems better than ChatGPT 4

2 min read

SomeOddCodeGuy

Dec 2 '23

Quick Start Guide To Converting Your Own GGUFs (including fp16)

6 min read

SomeOddCodeGuy

Nov 3 '23

Running old gglmv3 models in gpt4all

1 min read

SomeOddCodeGuy

Nov 1 '23

Clearing up confusion: GPT 3.5-Turbo may not be 20b after all

2 min read

SomeOddCodeGuy

Sep 15 '23

4090 Ahoy...

1 min read

SomeOddCodeGuy

Sep 14 '23

NTK Scaling and Llama 2

1 min read

SomeOddCodeGuy

Sep 4 '23

A Little About Local Models

2 min read

SomeOddCodeGuy

Aug 30 '23

Redownloading All The Old Models

1 min read

SomeOddCodeGuy

Jul 8 '23

If You Buy A Mac for LLMs, Don't Skimp on RAM

1 min read