Advice:
#TensorFlow is a safe bet for most projects. Not perfect but has huge community, wide usage. Maybe pair with high-level wrapper (#Keras, #Sonnet, etc)
I think #PyTorch is best for #research. However still new, there can be rough patches.
Use TensorFlow for one graph over many machines
✔️Consider #Caffe, #Caffe2, or TensorFlow for production deployment
✔️Consider TensorFlow or Caffe2 for #mobile
🖇source: http://cs231n.stanford.edu/slides/2017/cs231n_2017_lecture8.pdf
#deep_learning #framework
#TensorFlow is a safe bet for most projects. Not perfect but has huge community, wide usage. Maybe pair with high-level wrapper (#Keras, #Sonnet, etc)
I think #PyTorch is best for #research. However still new, there can be rough patches.
Use TensorFlow for one graph over many machines
✔️Consider #Caffe, #Caffe2, or TensorFlow for production deployment
✔️Consider TensorFlow or Caffe2 for #mobile
🖇source: http://cs231n.stanford.edu/slides/2017/cs231n_2017_lecture8.pdf
#deep_learning #framework