Loktar 🇺🇸

@loktar00

Building with local LLMs, retro PCs, and 30 years of making things. Christian. Dad. Veteran.

Michigan

Joined December 2022

Pinned
Loktar 🇺🇸
@loktar00
Jul 29
Qwen 3.6 27b Bf16 vs Q4 Not cherry picked first result from each, the prompt is from @ZenMagnets check out his pinned post it's a great test! I did throw Q4 a bone and gen twice more.. 2nd was also all white and the third was an error. Qwen 27b has a lot of life in it still!
00:00
Loktar 🇺🇸
@loktar00
47m
Probably my last Qwen 3.6 27b video outside of comparisons. Since switching to BF16 I've started to fully see what 27b has been capable all along These aren't perfect by any means but local.. but if a Jr. dev would have presented any of these in a portfolio I would be impressed
00:00
Loktar 🇺🇸
@loktar00
58m
Bah another finetuning miss. SO MANY lessons learned. Also side note, its already getting long in the tooth but this book is excellent. The Cranky Man's Guide to LoRA & QLoRA: Personal Lessons from a Thousand LLM Fine-Tuning Fails
Loktar 🇺🇸
@loktar00
1h
I'm back at 0% usage for Claude and Fable, ask me about this one simple trick that only cost $200!!
Loktar 🇺🇸
@loktar00
4h
Well.... this morning Microcenter increased prices for all 6000's... 3rd price increase in 7 days. I wouldn't be surprised if the Spark does see a mild bump soon at this point which is a bummer.