Skip to content

GitLab

Switch branch/tag
  1. 01 Apr, 2021 1 commit
  3. 31 Mar, 2021 2 commits
  5. 30 Mar, 2021 5 commits
  7. 29 Mar, 2021 6 commits
  9. 26 Mar, 2021 4 commits
  11. 25 Mar, 2021 5 commits
  13. 24 Mar, 2021 5 commits
  15. 23 Mar, 2021 8 commits
  17. 22 Mar, 2021 4 commits
    • Yeqing Li's avatar
      Internal change · 7bf19211
      Yeqing Li authored 
      PiperOrigin-RevId: 364433023
      7bf19211
    • A. Unique TensorFlower's avatar
      Internal change · b83ef168
      A. Unique TensorFlower authored 
      PiperOrigin-RevId: 364395434
      b83ef168
    • Poorva Potdar's avatar
      Internal change · 8cb5ac1e
      Poorva Potdar authored 
      PiperOrigin-RevId: 364378436
      8cb5ac1e
    • A. Unique TensorFlower's avatar
      Add a new option "recreate_iterator_for_each_eval" into StandardEvaluator. · 0e6f8848
      A. Unique TensorFlower authored 
      In current implementation of StandardEvaluator, the iterator of evaluate dataset will go back (actually create a new iterator) to the beginning when evaluate() is called every time.

In some case, the iterator need to go ahead forever. So add the new option "recreate_iterator_for_each_eval" to control this behavior.

In some case, the iterator creation is time consuming. For example, careting an iterator of distributed dataset for multi worker in eager mode. So it's better to make a repeat dataset instead of recreating an iterator.

If "recreate_iterator_for_each_eval" is True, then the iterator will go back; Otherwise, keep going ahead.

PiperOrigin-RevId: 364356169
      0e6f8848