All pull requests

Pull requests

• Document monarch profile (#5117)

  CLA Signedmeta-exported
  #5117·thomasywang opened Oct 22, 2026Contributor·

  ·
  1

• Add GPU metrics to distributed telemetry (#5114)

  CLA Signedmeta-exported
  #5114·thedavekwon opened Oct 22, 2026Contributor·

  ·
  1

• Document host and GPU metrics (#5113)

  CLA Signedmeta-exported
  #5113·thedavekwon opened Oct 22, 2026Contributor·

  ·
  1

• Release the GIL during telemetry scans (#5112)

  CLA Signedmeta-exported
  #5112·thedavekwon opened Oct 22, 2026Contributor·

  ·
  1

• Add Python helpers for telemetry socket ingest (#5111)

  CLA Signedmeta-exported
  #5111·thedavekwon opened Oct 22, 2026Contributor·

  ·
  1

• Sample host metrics into distributed telemetry (#5110)

  CLA Signedmeta-exported
  #5110·thedavekwon opened Oct 22, 2026Contributor·

  ·
  1

• Fix context_restorer double-consuming the context generator

  CLA Signed
  #5108·meilleure-lulu opened Oct 22, 2026·

  ·
  2

• one Handle state machine for receives and stream items

  CLA Signedmeta-exported
  #5106·shayne-fletcher opened Oct 21, 2026Contributor·

  ·
  1

• replace Future's _Complete state with a ready Handle

  CLA Signedmeta-exported
  #5105·shayne-fletcher opened Oct 21, 2026Contributor·

  ·
  1

• Add raw message actors and layer streaming as a protocol library

  CLA Signed
  #5081·dulinriley opened Oct 19, 2026Contributor·

  ·
  

• Attribute declined supervision failures to the declining actor (#5072)

  CLA Signedmeta-exported
  #5072·dulinriley opened Oct 17, 2026Contributor·

  ·
  1

  

• Add PBSJob, a PBS scheduler for monarch.job

  CLA Signed
  #5067·songhappy opened Oct 15, 2026·

  ·
  32

  

• fix Future.get() Tokio refusal text and warning location

  CLA Signedmeta-exported
  #5062·shayne-fletcher opened Oct 14, 2026Contributor·

  ·
  1

• keep logging and undeliverable test actors in-process

  CLA Signedmeta-exported
  #5019·shayne-fletcher opened Oct 7, 2026Contributor·

  ·
  1

• mesh admin: keep worker hosts visible for attached clients

  CLA Signedmeta-exported
  #5005·johnwhumphreys opened Oct 5, 2026Contributor·

  ·
  1

• Build FUSE RDMA wheel in CUDA builder image

  CLA Signedmeta-exported
  #5002·cpuhrsch opened Oct 2, 2026Contributor·

  ·
  1

• Fault injection

  CLA Signedmeta-exported
  #4978·thomasywang opened Oct 1, 2026Contributor·

  ·
  1

• Specify supervisors for chaos (#4977)

  CLA Signedmeta-exported
  #4977·thomasywang opened Oct 1, 2026Contributor·

  ·
  1

• Expose actor failure injection bindings to Python

  CLA Signedmeta-exported
  #4976·thomasywang opened Oct 1, 2026Contributor·

  ·

• Add ActorErrorKind::Injected

  CLA Signedmeta-exported
  #4975·thomasywang opened Oct 1, 2026Contributor·

  ·

• Allow user to define unique handlers for each DeliveryFailureKind (#4974)

  CLA Signedmeta-exported
  #4974·thomasywang opened Oct 1, 2026Contributor·

  ·
  1

• fix macOS CI failures

  CLA Signed
  #4949·dulinriley opened Sep 28, 2026Contributor·

  ·
  

• ROCM - Ionic NIC support

  ciflow/rocmCLA Signedmodule: rocm
  #4948·alugorey opened Sep 28, 2026Collaborator·

  ·

• Accept YAML Kubernetes exec credentials

  CLA Signedmeta-exported
  #4916·johnwhumphreys opened Sep 18, 2026Contributor·

  ·
  1

• Remove retained HostMeshRef from ProcMeshRef

  CLA Signedmeta-exportedmodule: rocm
  #4910·thomasywang opened Sep 17, 2026Contributor·

  ·
  1