HN
Hacker News
Top
New
Best
Ask
Show
Jobs
Comment by singpolyma3 | Hacker News Reader
Parent
Full thread
singpolyma3
·
The author already did and they were pretty good at it.
View on HN
dom96
·
On simple tasks, sure. Once a model reaches 100% I'll build a harder benchmark task set.
Reply on news.ycombinator.com