Hi,

"pelzflorian (Florian Pelz)" <[email protected]> skribis:

> Brendan Tildesley <[email protected]> writes:
>
>> To me, the "preferred source"
>> of these weights is the source code and dataset Intel used to generate
>> the weights, which I don't think are published.

I agree.

There is much debate about it because usually few entities have the
resources to train those neural networks on said data (I don’t know if
that’s the case for openimagedenoise).

> The Open-source Initiative cares about transparency where the training
> data is from. [1]

What the OSI did is also widely criticized:

  https://lwn.net/Articles/995159/

Ludo’.

Reply via email to