1. 18 Aug, 2023 1 commit
    • Xuechen Li's avatar
      support when num_heads is not divisible by world_size; resolves #459 (#461) · bb4cded1
      Xuechen Li authored
      * uneql rank.
      
      * trim.
      
      * enable passing in number of heads for each rank.
      
      * simplify.
      
      * simplify.
      
      * cleanup.
      
      * fix col parallel.
      
      * fix bug with row parallel.
      
      * fit out proj.
      
      * refac.
      
      * fix sharding logic.
      
      * refac sharding.
      
      * refac.
      
      * support multiple of.
      
      * make fn reuseable.
      
      * fix bug in dimensions.
      
      * scaffold.
      
      * test uneven heads.
      
      * fix test by adding barrier.
      
      * refac.
      
      * reuse code.
      
      * clean up.
      bb4cded1
  2. 17 Aug, 2023 1 commit
  3. 10 Aug, 2023 1 commit
  4. 26 Jul, 2023 2 commits
  5. 23 Jul, 2023 2 commits
  6. 22 Jul, 2023 1 commit
  7. 21 Jul, 2023 1 commit
  8. 17 Jul, 2023 1 commit
  9. 02 Jul, 2023 1 commit
  10. 02 Jun, 2023 1 commit
  11. 30 May, 2023 1 commit
  12. 06 May, 2023 1 commit
  13. 21 Apr, 2023 2 commits
  14. 19 Apr, 2023 1 commit
  15. 18 Apr, 2023 2 commits
  16. 14 Apr, 2023 1 commit
  17. 13 Apr, 2023 1 commit
  18. 31 Mar, 2023 1 commit
  19. 29 Mar, 2023 1 commit
  20. 22 Mar, 2023 1 commit
  21. 18 Jan, 2023 1 commit
  22. 16 Jan, 2023 3 commits
  23. 15 Jan, 2023 1 commit
  24. 08 Jan, 2023 1 commit
  25. 07 Jan, 2023 3 commits
  26. 04 Jan, 2023 1 commit
  27. 02 Jan, 2023 1 commit
  28. 01 Jan, 2023 1 commit
  29. 28 Dec, 2022 2 commits
  30. 25 Dec, 2022 2 commits