1. 19 Nov, 2025 1 commit
    • Kane's avatar
      Mlu590 deployment (#453) · fcc2a411
      Kane authored
      Feature:
          1. added mlu590 bfloat16, single-gpu and multi-gpus inference.
          2. added mlu590 int8 inference.
      fcc2a411
  2. 13 Nov, 2025 1 commit
  3. 24 Oct, 2025 2 commits
  4. 16 Oct, 2025 1 commit
  5. 29 Sep, 2025 2 commits
  6. 18 Sep, 2025 1 commit
  7. 03 Sep, 2025 1 commit
  8. 02 Sep, 2025 2 commits
  9. 01 Sep, 2025 2 commits
  10. 28 Aug, 2025 1 commit
  11. 27 Aug, 2025 2 commits
  12. 26 Aug, 2025 1 commit
  13. 25 Aug, 2025 1 commit
    • sandy's avatar
      Fix/wan2 2 vae encode api (#244) · 87343386
      sandy authored
      * bugfix:adapt to  5B dit model, derive attention_head_dim from config[dim]
      
      * [Fix] Wan2.2 Vae Encode refactor: drop args parameter and use self.cpu_offload
      87343386
  14. 20 Aug, 2025 2 commits
  15. 14 Aug, 2025 4 commits
  16. 12 Aug, 2025 1 commit
  17. 11 Aug, 2025 1 commit
  18. 09 Aug, 2025 1 commit
  19. 08 Aug, 2025 2 commits
  20. 05 Aug, 2025 2 commits
  21. 31 Jul, 2025 1 commit
  22. 30 Jul, 2025 2 commits
  23. 14 Jul, 2025 1 commit
  24. 11 Jun, 2025 1 commit
  25. 22 May, 2025 2 commits
  26. 14 May, 2025 1 commit
  27. 29 Apr, 2025 1 commit