CI (virtgpu)

Workflow runs for CI (virtgpu)

• ggml-cuda: HIP: use __builtin_amdgcn_perm for q1_0 unpacking (#29927)

  completed successfully:
  CI (virtgpu)#47:Commit9d1878apushedbymax-krasnyansky
  master

  Oct 17, 20264m 53s
  More actions for workflow run 47

• convert : fix Granite Switch block_count for MultiSwitch checkpoints …

  completed successfully:
  CI (virtgpu)#46:Commit59d5fadpushedbymax-krasnyansky
  master

  Oct 15, 20264m 37s
  More actions for workflow run 46

• rpc : turn GGML_RPC_DEBUG into a verbosity level and add logs (#29544)

  completed successfully:
  CI (virtgpu)#45:Commit5c8854fpushedbymax-krasnyansky
  master

  Oct 11, 20264m 37s
  More actions for workflow run 45

• sycl : Q5_K reorder-layout MMVQ and fused GLU (#29375)

  completed successfully:
  CI (virtgpu)#44:Commit3d65c90pushedbylhez
  master

  Oct 9, 20265m 1s
  More actions for workflow run 44

• qwen4exp: fix tests (#29819)

  completed successfully:
  CI (virtgpu)#43:Commit159c651pushedbymax-krasnyansky
  master

  Oct 2, 20264m 48s
  More actions for workflow run 43

• server: return HTTP 400 for invalid embedding requests (#29060)

  completed successfully:
  CI (virtgpu)#42:Commitd775ebfpushedbymax-krasnyansky
  master

  Oct 1, 20264m 45s
  More actions for workflow run 42

• hex-workqueue: fix race condition in seqn getting out of sync with id…

  completed successfully:
  CI (virtgpu)#41:Commit3aa0ce9pushedbylhez
  master

  Oct 1, 20264m 49s
  More actions for workflow run 41

• opencl: add bin kernel `kernel_gemm_noshuffle_q5_k_f32_32b_trans_ila_…

  completed successfully:
  CI (virtgpu)#40:Commita25c986pushedbymax-krasnyansky
  master

  Sep 25, 20264m 27s
  More actions for workflow run 40

• hexagon: add q5_k quant type support (#29123)

  completed successfully:
  CI (virtgpu)#39:Commit4de0926pushedbymax-krasnyansky
  master

  Sep 25, 20264m 22s
  More actions for workflow run 39

• metal : optimize sparse FA + clean-up (#29377)

  completed successfully:
  CI (virtgpu)#38:Commitcdc0642pushedbymax-krasnyansky
  master

  Sep 24, 20265m 0s
  More actions for workflow run 38

• hexagon: handle multi-sequence in concat_2d (#29344)

  completed successfully:
  CI (virtgpu)#37:Commita02c7f5pushedbymax-krasnyansky
  master

  Sep 24, 20264m 28s
  More actions for workflow run 37

• CUDA: tune FA for Gemma 4 on Ampere or newer (#29152)

  completed successfully:
  CI (virtgpu)#36:Commitce8caa6pushedbymax-krasnyansky
  master

  Sep 21, 20264m 39s
  More actions for workflow run 36

• hexagon: Support for K-Quants Q4_K and Q6_K (#28994)

  completed successfully:
  CI (virtgpu)#35:Commit1ec8188pushedbylhez
  master

  Sep 16, 20264m 34s
  More actions for workflow run 35

• opencl: extend the elementwise and data‐movement op coverage (#27633)

  completed successfully:
  CI (virtgpu)#34:Commit1548a24pushedbylhez
  master

  Sep 4, 20264m 44s
  More actions for workflow run 34

• common, server : enable preserve_reasoning kwarg by default, log its …

  completed successfully:
  CI (virtgpu)#33:Commite750b88pushedbymax-krasnyansky
  master

  Sep 2, 20264m 34s
  More actions for workflow run 33

• chat : add new template for DeepSeek V4 Flash 0731 (#26398)

  completed successfully:
  CI (virtgpu)#32:Commit0ef6e55pushedbytboinovski1
  master

  Aug 3, 20265m 8s
  More actions for workflow run 32

• ui: reduce per-token render cost when streaming (#26053)

  completed successfully:
  CI (virtgpu)#31:Commit555881epushedbylhez
  master

  Jul 25, 20265m 8s
  More actions for workflow run 31

• llama_dsv4: write only used rows in state (#25325)

  completed successfully:
  CI (virtgpu)#30:Commit91d2fc3pushedbytboinovski1
  master

  Jul 20, 20265m 21s
  More actions for workflow run 30

• opencl: disable FA and MoE weights repack to work around compiler iss…

  completed successfully:
  CI (virtgpu)#29:Commit3dafb58pushedbylhez
  master

  Jul 16, 20265m 10s
  More actions for workflow run 29

• opencl: exclude some moe kernels on Adreno a7x (#25698)

  completed successfully:
  CI (virtgpu)#28:Commit505b1edpushedbylhez
  master

  Jul 15, 20264m 59s
  More actions for workflow run 28

• vulkan: Use native e2m1 and e4m3 conversions for mxfp4/nvfp4 (#25338)

  completed successfully:
  CI (virtgpu)#27:Commite920c52pushedbylhez
  master

  Jul 13, 20265m 26s
  More actions for workflow run 27

• vulkan: disable FA mask_opt on GCN to improve performance (#24362)

  completed successfully:
  CI (virtgpu)#26:Commita646006pushedbylhez
  master

  Jul 8, 20265m 8s
  More actions for workflow run 26

• Add Q2_0 quantization: type definition and CPU backend (#24448)

  completed successfully:
  CI (virtgpu)#25:Commitbec4772pushedbylhez
  master

  Jul 7, 20265m 4s
  More actions for workflow run 25

• opencl: allow loading precompiled binary kernels from library (#23042)

  completed successfully:
  CI (virtgpu)#24:Commit4fc4ec5pushedbylhez
  master

  Jul 1, 20265m 18s
  More actions for workflow run 24

• Revert "sched : reintroduce less synchronizations during split comput…

  completed successfully:
  CI (virtgpu)#23:Commit86b9470pushedbylhez
  master

  Jun 30, 20265m 18s
  More actions for workflow run 23