[markeroon]

Imo the biggest issue (from memory) was that Tensorflow used a static computation graph. PyTorch was so much easier to work with.

[lostmsu]

I honestly think the static graph was much better in the similar way Vulkan/DX12 are better than OpenGL/DX9. It is harder to program, but gives more explicit control of important things. E.g. who would use PyTorch if they new that for optimal performance they'd need to record CUDA graphs?

[ianbutler]

Yup that was also definitely one among the many issues