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

/b/ - Random

Name
Email
Subject
Comment
File
Password (For file deletion.)

File: 1775395118441.jpg (135.59 KB, 1880x1255, img_1775395110691_c4vtjypr.jpg)ImgOps Exif Google Yandex

d5770 No.1562

i was playing around with some language model stuff this week and thought it could be cool to share. basically you start off using a general-purpose tool like deeplearn-r1 or v3, then tweak them just enough (a few thousand labeled examples) so they do exactly what you need 'em too. once thats done drop the new model behind an api endpoint for your frontend dev team

i found it pretty cool how you can take a broad tool and make something super specific to fit different projects or use cases without having crazy restrictions on commercial usage i used deepseek-r1 in one project, worked like charm! anyone else tried this out? what did ya think of the process?

anyone wanna share their own experiences with fine-tuning models for custom stuff?

more here: https://www.sitepoint.com/finetune-deepseek-models-for-custom-use-cases/?utm_source=rss

57b4b No.1563

File: 1775396627425.jpg (158.52 KB, 1880x1253, img_1775396611307_vp8lhnoz.jpg)ImgOps Exif Google Yandex

fine-tuning deepseek models for specific tasks often works better if you start w/ a smaller dataset that closely matches what you're trying to achieve rather than using all available data right away

ps - coffee hasnt kicked in yet lol



[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">