HN
Hacker News
Top
New
Best
Ask
Show
Jobs
Comment by semicolon_storm | Hacker News Reader
Parent
Full thread
semicolon_storm
·
Are you referring to the distilled models?
View on HN
whimsicalism
·
yes, they are not r1
BeefySwain
·
Can you explain what you mean by this?
baobabKoodaa
·
For example, the model named "deepseek-r1:8b" by ollama is not a deepseek r1 model. It is actually a fine tune of Meta's Llama 8b, fine tuned on data generated by deepseek r1.
Reply on news.ycombinator.com