All pull requests

Pull requests

• [RLlib] Exclude masked timesteps from GAE normalization

  community-contributionrllib
  #67663·Whning0513 opened Oct 31, 2026Contributor·

  ·
  1

  1

• [Data] Persist BlockMetadata.num_rows and batch ray.get in _calculate_blocks_rows

  community-contributiondata
  #67662·deep1blue opened Oct 31, 2026·1 task done·

  ·
  3

• [Data] Add fill_null expression method

  community-contributiondata
  #67661·surawut-jirasaktavee opened Oct 31, 2026·

  ·
  5

• [Data] Fix flaky test_torch_tensor_utils: raise Bazel size to medium

  #67660·AarryaSaraf opened Oct 31, 2026Contributor·

  ·

• [Data] Fix flaky test_data_iterator_checkpointer: wait for a readable checkpoint file

  #67659·AarryaSaraf opened Oct 31, 2026Contributor·

  ·

• [core][util] Fix ray.util.queue.Queue put/get with timeout=0

  community-contributioncore
  #67658·deep1blue opened Oct 31, 2026·4 of 8 tasks·

  ·

• [Data] Support Std and AbsMax in GPU hash aggregation

  community-contributiondata
  #67657·WangGuangxin opened Oct 31, 2026·

  ·
  6

• [Data] Push AND conjuncts on passthrough columns below a Project

  community-contributiondata
  #67656·sahil-mangla opened Oct 31, 2026·

  ·
  1

  4

• [data][llm] Split SGLang engine stage and processor into per-task classes

  community-contributiondata
  #67655·fortunecookiee opened Oct 31, 2026Contributor·

  ·
  4

• [Tune] Preserve errored trials when importing evaluated results

  community-contributiontune
  #67654·MakerFlyFly opened Oct 31, 2026·

  ·
  1

  1

• [Data][CI] Run the Ray Data pyrefly check in pre-commit

  datadevprod
  #67652·goutamvenkat-anyscale opened Oct 31, 2026Contributor·

  ·
  1

• [Train] Shut down failed worker group before sizing replacement

  community-contributiontrain
  #67651·ryanaoleary opened Oct 31, 2026Contributor·

  ·
  5

• [Data] Estimate PyArrow's peak memory decoding a Parquet row group from its footer [5/5]

  #67650·AarryaSaraf opened Oct 31, 2026Contributor·

  ·
  1

• [Data] Log the conflicting column and types on schema unification error

  datago
  #67647·ayushk7102 opened Oct 31, 2026Contributor·

  ·
  8

• [Data] Drain Iterator to Fix Flakiness in Rate-Based Cluster Autoscaler Test

  datago
  #67646·rayhhome opened Oct 31, 2026Contributor·

  ·
  3

  

• [Data] Skip the S3 HEAD request on each Parquet file open by passing listed file sizes

  datago
  #67645·AarryaSaraf opened Oct 31, 2026Contributor·

  ·
  5

• Bump follow-redirects in /python/ray/dashboard/client

  coredependenciesjavascript
  #67644·dependabot[bot] opened Oct 30, 2026Bot·

  ·
  1

• [Data] Size Parquet read tasks by their decoded Arrow bytes in DSv2 [4/5]

  data
  #67642·AarryaSaraf opened Oct 30, 2026Contributor·

  ·
  1

• [Data] Size Ray's read options, schema pin and added columns in the Parquet decoded-size estimate [3/5]

  data
  #67641·AarryaSaraf opened Oct 30, 2026Contributor·

  ·
  4

• [Data] Estimate a Parquet row group's decoded Arrow bytes from its footer [2/5]

  datapython
  #67640·AarryaSaraf opened Oct 30, 2026Contributor·

  ·
  1

• [doc] Limit the in-tree notebook page to maintaining existing examples

  coredocsgo
  #67639·dstrodtman opened Oct 30, 2026Contributor·

  ·
  2

• [ci] Check docstring and Returns: presence on new public APIs

  #67638·dstrodtman opened Oct 30, 2026Contributor·

  ·
  1

• Add perf metrics for 2.60.0

  community-contributioncorerelease-test
  #67637·khluu opened Oct 30, 2026Contributor·

  ·
  3

  

• [core] Type-annotate untyped Core API signatures so their parameters can be documented

  corego
  #67636·dstrodtman opened Oct 30, 2026Contributor·

  ·
  1

• [core] Add missing docstrings, Args, and Returns to ray.util, ray.experimental, and the actor APIs

  coredocstring-onlygo
  #67635·dstrodtman opened Oct 30, 2026Contributor·

  ·
  1