Pull requests: intel/intel-xpu-backend-for-triton

19 Open 1,433 Closed

Author

Filter by author

Loading

author:

Filter by this user

Label

Filter by label

Loading

Projects

Filter by project

Loading

Milestones

Filter by milestone

Loading

Reviews

Filter by reviews

No reviews Review required Approved review Changes requested

Assignee

Filter by who’s assigned

Assigned to nobody Loading

assignee:

Filter by this user

Sort

Sort by

Newest Oldest Most commented Least commented Recently updated Least recently updated Best match

Most reactions

👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Support causal flash attention

#2425 opened Oct 4, 2024 by jopperm Loading…

1

1

Add dotOperandDpasToLinearLayout (Tensor B)

#2422 opened Oct 3, 2024 by whitneywhtsang Loading…

1

Use block load attribute to remove duplicate logic from MaterializeBlockPointer pass

#2420 opened Oct 3, 2024 by alexbaden • Draft

1

1

Feature/benchmark improvements enhancement New feature or request

#2415 opened Oct 2, 2024 by ZzEeKkAa • Draft

1

4

Add new skip list for Xe2 based GPUs such as LNL

#2405 opened Oct 2, 2024 by arunjose696 Loading…

1

2

Return n_regs for binaries compiled explicitly with a register size mode option

#2391 opened Oct 1, 2024 by alexbaden • Draft

1

Add XeTLA FA backward implementation to benchmark

#2367 opened Sep 27, 2024 by ESI-SYD Loading…

1

2

[intel] Remove RewriteTensorPointer pass

#2359 opened Sep 26, 2024 by whitneywhtsang • Draft

2

5

[testing] Enable generating cached native code by default

#2350 opened Sep 25, 2024 by alexbaden • Draft

1

[CI] show less verbose log and give failed summary for each commit id

#2319 opened Sep 24, 2024 by AshburnLee Loading…

1

4

Increase cache size

#2264 opened Sep 16, 2024 by anmyachev • Draft

1

2

1471: Generic SPIR-V Runner

#2258 opened Sep 16, 2024 by kballeda Loading…

1

73

Increase warmup and rep for FA benchmark

#2256 opened Sep 16, 2024 by anmyachev Loading…

1

18

Remove usage of torch.autograd.profiler_legacy for benchmarks

#2149 opened Sep 6, 2024 by anmyachev • Draft

1

6

Remove import intel_extension_for_pytorch from benchmarks

#2147 opened Sep 6, 2024 by anmyachev • Draft

1

2

Support causal masking in FlashAttention

#1947 opened Aug 20, 2024 by jopperm • Draft

1

1

[WIP][scripts] Collect env diagnostic script

#1926 opened Aug 19, 2024 by kwasd • Draft

1

Build benchmarks with upstream pytorch

#1905 opened Aug 16, 2024 by ZzEeKkAa • Draft

1

30

Use select for masked single element load

#1770 opened Aug 2, 2024 by alexbaden • Draft

1

4