Datasets:
Important - Copyright Question
#22
by Enderchef - opened
Is the data in fineweb-edu copyrighted? If so, can you make a version with not-copyrighted material?
Hmmm
Is AI trained with copyright material
I wonder.
https://www.npr.org/2025/09/05/nx-s1-5529404/anthropic-settlement-authors-copyright-ai
lol.
That was a long time ago when I was new to ai
Lol
Enderchef changed discussion status to closed
Oh yeah, back in 2025 π€£
Didnt see the date
You planning to train your small model(Haiku, I think) on this?
Lol π
Yes actually :D
Next haiku is 10B tokens from this dataset.