3 ms·

Anyone knows why Andrej's team chooses PyTorch (as oppose to say TensorFlow?)

• ankeshanand 7y ago
  https://twitter.com/karpathy/status/868178954032513024 https://twitter.com/karpathy/status/868178954032513024
• tigershark 7y ago
  Not at expert, but as far as I understood PyTorch is much better to build new models, while with tensorflow it’s easier to assemble the predefined blocks. Source: somewhere in the motivations on why Fast.Ai courses switched to PyTorch for the second edition.
• jeffshek 7y ago
  Some potential reasons: - TensorFlow is great at deployment, but not the easiest to code. PyTorch isn't frequently used in production until recently. - If you have the resources for great AI engineers and researchers, your team will be good enough to build and deploy both frameworks. - Preference toward the easier framework your tech leads prefer. - Lots of new academic research is coming in PyTorch - TensorFlow is undergoing a massive change from 1.1x to 2.0; if you choose TensorFlow, write on 1.1x just to then refactor to TF 2.0? Or write on TF 2.0 now and deal with all new edge cases? Or write in PyTorch (easier) but handle the more difficult deployment process. - ML code quickly rots. Bad PyTorch code is just bad Python code. Bad TensorFlow code can be a nightmare to debug. - PyTorch's eager execution makes coding NNs much easier to prototype and build.
• m0zg 7y ago
  Because PyTorch literally triples researcher productivity. Imagine a deep learning framework which you can actually debug when something goes wrong and which you don't have to fight every step of the way to do even simple things. That's PyTorch.