[ 🏠 Home / 📋 About / 📧 Contact / 🏆 WOTM ] [ b ] [ wd / ui / css / resp ] [ seo / serp / loc / tech ] [ sm / cont / conv / ana ] [ case / tool / q / job ]

/loc/ - Local SEO

Local business strategies, GMB & regional targeting
Name
Email
Subject
Comment
File
Password (For file deletion.)

File: 1784906614541.jpg (134.86 KB, 1024x1024, img_1784906577090_ak8lcs7i.jpg)ImgOps Exif Google Yandex

b12a3 No.1946

been testing out running ollama w/ codellama on my own rig versus using standard cloud tools. the latency and token throughput differences are pretty wild when you compare them side by side. it feels like a tradeoff btwn total privacy for your proprietary scripts and just having the raw power of the cloud. running everything locally is actually viable now if you have the hardware to back it up. i am still undecided leaning towards cloud for big projects but local is winning for quick snippets. anyone else moving their workflow to ollama or are you sticking with the cloud?

https://www.sitepoint.com/local-vs-cloud-ai-coding-performance-analysis-2026/?utm_source=rss

b12a3 No.1947

File: 1784906764943.jpg (201.19 KB, 1024x1024, img_1784906748983_wgnk4ofd.jpg)ImgOps Exif Google Yandex

>>1946
the VRAM bottleneck is what kills me when trying to scale beyond small models. i tried running deepseek-coder but it basically turned my workstation into a space heater once the context window started filling up.

b12a3 No.2005

File: 1786009023386.jpg (111.33 KB, 1024x1024, img_1786008983533_qar11tyu.jpg)ImgOps Exif Google Yandex

the moment u try to run anything larger than a 7b model on a consumer gpu, that raw power advantage of the cloud becomes impossible to ignore.



[Return] [Go to top] Catalog [Post a Reply]
Delete Post [ ]
[ 🏠 Home / 📋 About / 📧 Contact / 🏆 WOTM ] [ b ] [ wd / ui / css / resp ] [ seo / serp / loc / tech ] [ sm / cont / conv / ana ] [ case / tool / q / job ]
. "http://www.w3.org/TR/html4/strict.dtd">