François CholletVerified account

Interesting result - VGG 2x faster on @pytorch vs keras + @tensorflow https://medium.com/@vishnuvig/transfer-learning-using-pytorch-part-2-9c5b18e15551 … Seems to mainly be due to preprocessing time

or using TFs input queues, gives me around 3xspeedup. But not as easy to use as Keras image datagenerato..

For the model where I compared the result , I did not use any multiprocessing capability. Pytorch performance of 2x is without multiprocess

I tried this on my system, up'd to Keras 2. pytorch: 80s/epoch. keras 2: 93s/epoch. 1 worker each. w/o flip/normalize for fairer comparison

You can get even better performance by removing the overhead of feed_dict. Simply use a TF queue and build your Keras model on top of that.