Martins Udris's picture

Building on HF

6 3 2

Martins Udris PRO

martinsu

·

AI & ML interests

None yet

Recent Activity

upvoted an article 17 days ago

How We Built a Semantic Highlight Model To Save Token Cost for RAG

new activity about 1 month ago

martinsu/tildeopen-30b-mu-instruct:Use case

replied to etemiz's post about 2 months ago

I realized when I ask longer answers to my questions, the models sometimes produce completely opposite answer. What could be the reason? I do mostly CPT. Should I convert my dataset to SFT and give longer reasonings too for it to have integrity? Example: Is the yolk of an egg more beneficial or the white? Answer in 100 words. Answer: Yolk is more beneficial because .......... Example: Is the yolk of an egg more beneficial or the white? Answer in 500 words. Answer: White is more beneficial because .......... Edit: These happen in temp = 0.0

View all activity

Organizations

martinsu 's models 1

martinsu/tildeopen-30b-mu-instruct

Text Generation • 31B • Updated Dec 6, 2025 • 40 • 4

martinsu (Martins Udris)

Martins Udris's picture

Building on HF

6 3 2

Martins Udris PRO

martinsu

·

AI & ML interests

None yet

Recent Activity

upvoted an article 17 days ago

How We Built a Semantic Highlight Model To Save Token Cost for RAG

new activity about 1 month ago

martinsu/tildeopen-30b-mu-instruct:Use case

replied to etemiz's post about 2 months ago

I realized when I ask longer answers to my questions, the models sometimes produce completely opposite answer. What could be the reason? I do mostly CPT. Should I convert my dataset to SFT and give longer reasonings too for it to have integrity? Example: Is the yolk of an egg more beneficial or the white? Answer in 100 words. Answer: Yolk is more beneficial because .......... Example: Is the yolk of an egg more beneficial or the white? Answer in 500 words. Answer: White is more beneficial because .......... Edit: These happen in temp = 0.0

View all activity

Organizations

martinsu 's models 1

martinsu/tildeopen-30b-mu-instruct

Text Generation • 31B • Updated Dec 6, 2025 • 40 • 4