Server (self-hosted)

Workflow runs for Server (self-hosted)

• webgpu : fix handling of infinity values during ARGSORT and TOP_K (#2…

  cancelled:
  Server (self-hosted)#31:Commit5ea87ddpushedbykhosravipasha
  master

  Aug 25, 20261d 0h 0m 2s
  More actions for workflow run 31

• mtmd: add lanczos resize method [no release] (#26341)

  cancelled:
  Server (self-hosted)#30:Commit5f55650pushedbykhosravipasha
  master

  Jul 30, 20261d 0h 0m 2s
  More actions for workflow run 30

• llama: move suppress_tokens handling to common/sampling (#26276)

  cancelled:
  Server (self-hosted)#29:Commitafeebe1pushedbykhosravipasha
  master

  Jul 29, 20261d 0h 0m 3s
  More actions for workflow run 29

• contrib: allow all AI-generated code in general (#26012)

  cancelled:
  Server (self-hosted)#28:Commit4310aa4pushedbykhosravipasha
  master

  Jul 22, 20261d 0h 0m 2s
  More actions for workflow run 28

• opencl: do not use clCreateBufferWithProperties when targeting CL 2…

  cancelled:
  Server (self-hosted)#27:Commit12127depushedbykhosravipasha
  master

  Jul 15, 20261d 0h 0m 2s
  More actions for workflow run 27

• metal : add Q2_0 support (#25419)

  cancelled:
  Server (self-hosted)#26:Commit14d3ba4pushedbykhosravipasha
  master

  Jul 14, 20261d 0h 0m 1s
  More actions for workflow run 26

• CUDA: refactor MMQ kernel configuration (#24127)

  cancelled:
  Server (self-hosted)#25:Commit6eddde0pushedbykhosravipasha
  master

  Jul 13, 20261d 0h 0m 1s
  More actions for workflow run 25

• ggml-webgpu: tune subgroup split (d_split) in flash_attn_vec (#25418)

  cancelled:
  Server (self-hosted)#24:Commit32e41fapushedbykhosravipasha
  master

  Jul 9, 20261d 0h 0m 1s
  More actions for workflow run 24

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

  cancelled:
  Server (self-hosted)#23:Commita646006pushedbykhosravipasha
  master

  Jul 8, 20261d 0h 0m 2s
  More actions for workflow run 23

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

  cancelled:
  Server (self-hosted)#22:Commitbec4772pushedbykhosravipasha
  master

  Jul 7, 20261d 0h 0m 5s
  More actions for workflow run 22

• Server (self-hosted)

  cancelled:
  Server (self-hosted)#21:Commit391c775pushedbykhosravipasha
  master

  Jul 7, 20261d 0h 0m 2s
  More actions for workflow run 21

• Server (self-hosted)

  cancelled:
  Server (self-hosted)#20:Commit78df1fcpushedbykhosravipasha
  master

  Jul 6, 20261d 0h 0m 2s
  More actions for workflow run 20

• Hexagon: Process M-tail rows on HMX instead of HVX (#22724)

  cancelled:
  Server (self-hosted)#19:Commitbbeb89dpushedbykhosravipasha
  master

  May 5, 20261d 0h 0m 2s
  More actions for workflow run 19

• pr2wt : symlink .pi (#22386)

  cancelled:
  Server (self-hosted)#18:Commitf535774pushedbykhosravipasha
  master

  Apr 26, 20261d 0h 0m 1s
  More actions for workflow run 18

• chat: fix parallel_tool_calls default setting based on model capabili…

  cancelled:
  Server (self-hosted)#17:Commit8bccdbbpushedbykhosravipasha
  master

  Apr 22, 20261d 0h 0m 2s
  More actions for workflow run 17

• server: Expose media_tag on /props endpoint. (#22028)

  cancelled:
  Server (self-hosted)#16:Commit9e5647apushedbykhosravipasha
  master

  Apr 18, 20261d 0h 0m 1s
  More actions for workflow run 16

• metal: Implement ROLL op (#21946)

  cancelled:
  Server (self-hosted)#15:Commitae2d348pushedbykhosravipasha
  master

  Apr 16, 20261d 0h 0m 2s
  More actions for workflow run 15

• cuda: Q1_0 initial backend (#21629)

  cancelled:
  Server (self-hosted)#14:Commit7e72b38pushedbykhosravipasha
  master

  Apr 15, 20261d 0h 0m 2s
  More actions for workflow run 14

• docs: listing qwen3-asr and qwen3-omni as supported (#21857)

  cancelled:
  Server (self-hosted)#13:Commite974923pushedbykhosravipasha
  master

  Apr 13, 20261d 0h 0m 1s
  More actions for workflow run 13

• convert : force f16 or f32 on step3-vl conv weights (#21646)

  cancelled:
  Server (self-hosted)#12:Commit1e9d771pushedbykhosravipasha
  master

  Apr 12, 20261d 0h 0m 1s
  More actions for workflow run 12

• CUDA: skip compilation of superfluous FA kernels (#21768)

  cancelled:
  Server (self-hosted)#11:Commitff5ef82pushedbykhosravipasha
  master

  Apr 12, 20261d 0h 0m 1s
  More actions for workflow run 11

• hexagon: add support for linux on snapdragon (#21707)

  cancelled:
  Server (self-hosted)#10:Commit81069a8pushedbykhosravipasha
  master

  Apr 10, 20261d 0h 0m 2s
  More actions for workflow run 10

• vulkan: Support Q1_0 (#21539)

  cancelled:
  Server (self-hosted)#9:Commit7b69125pushedbykhosravipasha
  master

  Apr 10, 20261d 0h 0m 2s
  More actions for workflow run 9

• CUDA: also store node->src->data ptrs for equality check (#21635)

  cancelled:
  Server (self-hosted)#8:Commitd12cc3dpushedbykhosravipasha
  master

  Apr 8, 20261d 0h 0m 1s
  More actions for workflow run 8

• ggml-webgpu: Add the support of MUL_MAT_ID (#21147)

  cancelled:
  Server (self-hosted)#7:Commitd0a6dfepushedbykhosravipasha
  master

  Apr 6, 20261d 0h 0m 2s
  More actions for workflow run 7