3 ms·
I'm not up to speed with GPLv3 vs GPLv2, but with GPLv2 you could technically develop the smallest possible functionality around the library, and make that part
by dguaraglia 8y ago
I'm not up to speed with GPLv3 vs GPLv2, but with GPLv2 you could technically develop the smallest possible functionality around the library, and make that part open source. At the end of the day, with any ML product it's gathering and labeling the training data that is the biggest differentiator.