1. 30 Mar, 2022 1 commit
  2. 03 Mar, 2022 1 commit
  3. 02 Mar, 2022 1 commit
  4. 15 Feb, 2022 3 commits
  5. 11 Feb, 2022 1 commit
  6. 01 Feb, 2022 3 commits
  7. 31 Jan, 2022 1 commit
  8. 29 Jan, 2022 1 commit
  9. 28 Jan, 2022 2 commits
  10. 27 Jan, 2022 1 commit
  11. 26 Jan, 2022 1 commit
  12. 24 Jan, 2022 2 commits
  13. 22 Jan, 2022 1 commit
  14. 19 Jan, 2022 1 commit
  15. 12 Jan, 2022 1 commit
  16. 21 Dec, 2021 1 commit
  17. 17 Dec, 2021 1 commit
  18. 02 Dec, 2021 1 commit
  19. 26 Nov, 2021 1 commit
  20. 24 Nov, 2021 1 commit
  21. 01 Oct, 2021 1 commit
  22. 20 Sep, 2021 1 commit
  23. 19 Aug, 2021 1 commit
  24. 18 Aug, 2021 2 commits
  25. 17 Aug, 2021 2 commits
  26. 16 Aug, 2021 2 commits
  27. 30 Jul, 2021 1 commit
    • Deepak Narayanan's avatar
      Support for pipeline parallelism in T5 model · 46c74b4c
      Deepak Narayanan authored
      - Accumulate encoder hidden state gradient to handle skip connection
      - Correctly compute the number of layers in encoder / decoder for T5 model
      - Ensure e weights are initialized the same way in embeddings
      - Synchronize embedding gradients across encoder and decoder for T5 model
      - Support for checkpoint loading and saving
      46c74b4c
  28. 21 Jul, 2021 1 commit
  29. 12 Jul, 2021 1 commit
  30. 05 Jul, 2021 2 commits