-1

A model is trained with GPU and result is saved by checkpoint file. The saved checkpoint file can be run by cpu-tensorflow? If not, can convert the saved checkpoint file so as to run model in cpu-tensorflow?

glennsl
  • 28,186
  • 12
  • 57
  • 75
liu bob
  • 21
  • 1
  • 7
    Possible duplicate of [Can a model trained on gpu used on cpu for inference and vice versa?](http://stackoverflow.com/questions/40980035/can-a-model-trained-on-gpu-used-on-cpu-for-inference-and-vice-versa) – kafman Mar 22 '17 at 13:19

1 Answers1

0

Yes! It normally can!

The exception is with tf.device('gpu:0') statements. If you do not have them in your code you are good to go!

Good luck!

rmeertens
  • 4,383
  • 3
  • 17
  • 42
  • 1
    If you do have device assignments in your code, you can remove them from the saved GraphDef .pb file with the remove_device rewriting rule: https://github.com/tensorflow/tensorflow/tree/master/tensorflow/tools/graph_transforms/#remove_device – Pete Warden Mar 22 '17 at 15:50