Open marsggbo opened 5 years ago
num_shards is no longer necessary in tf 1.14 and 1.13. It can be auto-derived.
Thanks for your quick reply @saberkun . I try again by removing the config of num_shards, it only use 8 cores instead of 32 cores.
Here is the TPU I created:
I have another question, i.e. does the code of AmoebaNet contains the process of searching or only the process of training?
Cloud TPU team are actively look into this.
For architecture search, the answer is no. In current repo, there is no RL-based search algorithm included.
@saberkun Is there any other open source code of NAS that can run on TPU? I want to compare the performance of TPU and GPU in terms of speed and accuracy, etc.
@marsggbo You might be interested in http://github.com/dstamoulis/single-path-nas.
@marsggbo Hi, do you mean other nets from architecture search? We have EfficientNet and MnasNet family inside this repo. I believe most imagenet classification problems can be easily implemented with these TPUEstimator framework. I could not recommend any search algorithm implementation as I did not look into them.
I'm running the code of AmoebaNet using a TPUv2-32, but it can only find 8 TPU cores,and raises the following errors:
What's the reason and hot to solve this problem?