Closed vincetom closed 6 years ago
I am using 4 cpu(Intel Xeon E5-2682v4), 60g memory, 16g GPU( Nvidia-p100).
you've better to use GPU as there are millions of training data & very long document, and cnn/rnn are computation expensive,。
for cloud gpu, please check: https://www.aliyun.com/product/ecs/gpu
great job! but when I use your code to train the 'big_data' (1700k),cpu always occupied 100%,then I cannot connect to my server.