1. 07 May, 2021 1 commit
  2. 16 Mar, 2021 1 commit
  3. 02 Mar, 2021 1 commit
  4. 01 Mar, 2021 2 commits
  5. 12 Oct, 2020 1 commit
  6. 22 Sep, 2020 1 commit
  7. 12 Aug, 2020 1 commit
  8. 10 Aug, 2020 1 commit
  9. 16 Jul, 2020 1 commit
  10. 15 Jul, 2020 1 commit
  11. 14 Jul, 2020 2 commits
  12. 13 Jul, 2020 4 commits
  13. 01 May, 2020 1 commit
  14. 21 Apr, 2020 1 commit
  15. 14 Apr, 2020 2 commits
  16. 13 Apr, 2020 1 commit
  17. 12 Apr, 2020 1 commit
  18. 11 Apr, 2020 1 commit
  19. 09 Apr, 2020 1 commit
  20. 03 Apr, 2020 1 commit
  21. 16 Mar, 2020 1 commit
  22. 13 Mar, 2020 1 commit
  23. 11 Mar, 2020 1 commit
  24. 09 Mar, 2020 1 commit
  25. 03 Mar, 2020 1 commit
  26. 06 Feb, 2020 1 commit
  27. 01 Feb, 2020 1 commit
  28. 28 Jan, 2020 1 commit
  29. 14 Dec, 2019 1 commit
  30. 27 Nov, 2019 1 commit
  31. 25 Nov, 2019 1 commit
    • Sai Ganesh Bandiatmakuri's avatar
      Inject enable_runtime_flags into benchmarks. · bcce419a
      Sai Ganesh Bandiatmakuri authored
      This will help general debugging by enabling custom execution with  --benchmark_method_steps.
      
      E.g --benchmark_method_steps=train_steps=7 will run the benchmark for only 7 steps without modifying benchmark code.
      
      PiperOrigin-RevId: 282396875
      bcce419a
  32. 01 Nov, 2019 1 commit
    • Reed Wanderman-Milne's avatar
      Support fp16 using tf.keras.mixed_precision in CTL resnet. · ffa522ea
      Reed Wanderman-Milne authored
      To test, I ran the following command:
      
      python resnet_ctl_imagenet_main.py --batch_size=2048 --data_dir ~/imagenet --datasets_num_private_threads=14 --epochs_between_evals=10 --model_dir ~/tmp_model_dir --clean --num_gpus=8 --train_epochs=90 --dtype=fp16
      
      I got 76.15% final evaluation accuracy.
      
      PiperOrigin-RevId: 278010061
      ffa522ea
  33. 18 Oct, 2019 1 commit
  34. 17 Oct, 2019 1 commit