Build and Publish GPU Image

Workflow runs for Build and Publish GPU Image

• fix(ci): resolve test-client image owner from github.repository_owner

  completed successfully:
  Build and Publish GPU Image#13:Commit6466a54pushedbybrandonm
  master

  Jul 15, 20261m 3s
  More actions for workflow run 13

• fix(ci): resolve test-client image owner from github.repository_owner

  completed successfully:
  Build and Publish GPU Image#12:Commit6466a54pushedbybrandonm
  feature/model-ttl

  Jul 15, 202640m 37s
  More actions for workflow run 12

• fix(inference): drain the worker pipe when a generation is abandoned

  completed successfully:
  Build and Publish GPU Image#11:Commitf3a88b4pushedbybrandonm
  master

  Jul 15, 202624m 9s
  More actions for workflow run 11

• fix(inference): drain the worker pipe when a generation is abandoned

  completed successfully:
  Build and Publish GPU Image#10:Commitf3a88b4pushedbybrandonm
  feature/model-ttl

  Jul 14, 202625m 24s
  More actions for workflow run 10

• ci(gpu): build cu128 (Blackwell) as the default GPU image

  completed successfully:
  Build and Publish GPU Image#9:Commit25205ffpushedbybrandonm
  feature/model-ttl

  Jun 20, 202641m 0s
  More actions for workflow run 9

• ci(cpu): add build-cpu workflow to publish CPU image to GHCR

  completed successfully:
  Build and Publish GPU Image#8:Commit29168c9pushedbybrandonm
  feature/model-ttl

  Jun 16, 202622m 34s
  More actions for workflow run 8

• ci(gpu): point build-gpu workflow at Dockerfile.optimized

  completed successfully:
  Build and Publish GPU Image#7:Commit0e4189apushedbybrandonm
  feature/model-ttl

  Jun 9, 202621m 10s
  More actions for workflow run 7

• Merge upstream/master (v0.5.0) into feature/model-ttl

  failed:
  Build and Publish GPU Image#6:Commit1d1761bpushedbybrandonm
  feature/model-ttl

  Jun 8, 202651s
  More actions for workflow run 6

• fix: warmup endpoint resets TTL timer

  completed successfully:
  Build and Publish GPU Image#5:Commit9dac681pushedbybrandonm
  feature/model-ttl

  Apr 8, 20269m 22s
  More actions for workflow run 5

• feat: add POST /v1/warmup endpoint for pre-spawning worker

  completed successfully:
  Build and Publish GPU Image#4:Commitc056e18pushedbybrandonm
  feature/model-ttl

  Apr 2, 20267m 54s
  More actions for workflow run 4

• feat: subprocess isolation for full GPU memory release

  completed successfully:
  Build and Publish GPU Image#3:Commit36929d5pushedbybrandonm
  feature/model-ttl

  Apr 2, 202612m 11s
  More actions for workflow run 3

• fix: gc.collect() before empty_cache to actually free GPU memory

  completed successfully:
  Build and Publish GPU Image#2:Commit55bd746pushedbybrandonm
  feature/model-ttl

  Apr 2, 202614m 36s
  More actions for workflow run 2

• feat: add MODEL_TTL for idle GPU unload

  completed successfully:
  Build and Publish GPU Image#1:Commit72662e0pushedbybrandonm
  feature/model-ttl

  Apr 2, 202646m 12s
  More actions for workflow run 1