1. 28 Mar, 2024 17 commits
  2. 19 Mar, 2024 1 commit
  3. 15 Mar, 2024 3 commits
  4. 14 Mar, 2024 2 commits
  5. 02 Mar, 2024 2 commits
  6. 21 Feb, 2024 4 commits
  7. 20 Feb, 2024 1 commit
  8. 18 Feb, 2024 1 commit
    • Qubitium's avatar
      Optimize compile to 1: avoid oom 2: minimize swap usage 3: avoid threads... · f45bbb4c
      Qubitium authored
      Optimize compile to 1: avoid oom 2: minimize swap usage 3: avoid threads starvation when letting ninja decide how many workers to spawn or manual MAX_JOBS "guesses". Logic is to take the min value of MAX_JOBS auto-calculated by two metrics: 1: cpu cores 2: free memory. This should allow flash-attn to compile close to the most efficient manner under any consumer/server env. (#832)
      
      f45bbb4c
  9. 10 Feb, 2024 4 commits
  10. 08 Feb, 2024 1 commit
  11. 31 Jan, 2024 3 commits
  12. 30 Jan, 2024 1 commit