1. 23 Mar, 2022 2 commits
  2. 22 Mar, 2022 3 commits
  3. 21 Mar, 2022 1 commit
  4. 14 Mar, 2022 1 commit
  5. 09 Mar, 2022 2 commits
  6. 02 Mar, 2022 1 commit
  7. 01 Mar, 2022 2 commits
  8. 28 Feb, 2022 1 commit
  9. 18 Feb, 2022 1 commit
  10. 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
  11. 04 Mar, 2021 2 commits
  12. 23 Feb, 2021 1 commit
  13. 26 Jan, 2021 1 commit
  14. 06 Jan, 2021 2 commits
  15. 05 Jan, 2021 1 commit
  16. 30 Dec, 2020 1 commit