10 Pull requests merged by 7 people

• Add log sigmoid

  #408 merged Feb 12, 2025

• [Metax] Add metax vdot heuristics config

  #439 merged Feb 12, 2025

• [Bugfix] skip quantile tests on triton lower than 3.0 and cummin benchmark

  #443 merged Feb 12, 2025

• MSELoss

  #435 merged Feb 12, 2025

• [Iluvatar] Add vdot_heur_block_size

  #440 merged Feb 12, 2025

• [Benchmark] skip conv2d perf test

  #441 merged Feb 11, 2025

• Fix a Setup error

  #437 merged Feb 8, 2025

• add backward of conv2d

  #365 merged Feb 7, 2025

• Add quantile op

  #287 merged Feb 6, 2025

• [Test] upcast cfloat to complex128 as reference

  #436 merged Feb 6, 2025

2 Pull requests opened by 2 people

• fix bugs on database search

  #438 opened Feb 8, 2025

• update some kernel names

  #444 opened Feb 12, 2025

2 Issues closed by 2 people

• Code Contribution: [Hard] [Operator Development] mse_loss

  #396 closed Feb 12, 2025

• Code Contribution: 【Lv1】【Operator Development】quantile

  #251 closed Feb 6, 2025

1 Issue opened by 1 person

• backward pass of flash attention in triton

  #442 opened Feb 11, 2025

8 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• Add_elu

  #415 commented on Feb 11, 2025 • 8 new comments

• [Operator] Add Conv3d forward function

  #412 commented on Feb 11, 2025 • 7 new comments

• Add index_put op

  #429 commented on Feb 7, 2025 • 4 new comments

• Add Op dot

  #430 commented on Feb 7, 2025 • 1 new comment

• Enhance MM Kernel Performance and Coverage for Specific Input Scenarios

  #405 commented on Feb 6, 2025 • 0 new comments

• [fix] rewrite do_bench with fixed warmup/rep runs mode

  #411 commented on Feb 7, 2025 • 0 new comments

• Backward register

  #423 commented on Feb 11, 2025 • 0 new comments

• Add Op Diff

  #428 commented on Feb 10, 2025 • 0 new comments