contribution/compass
ray-project/ray

Ray

114 signals · 115 observation events

Open repository ↗

Ray is an AI compute engine. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.

43.5K stars7.9K forksPythonApache-2.0data-sciencedeep-learningdeploymentdistributedhyperparameter-optimizationhyperparameter-searchlarge-language-modelsllmkeyword: Ray distributed computingkeyword: Ray framework
PROJECT NEWS

Release, roadmap, and discussion

All news →
ray-project/ray

Ray

AI Infrastructure
Latest stable

Ray-2.57.0

ray-2.57.0
  • Ray Data: In this Ray release, we've enabled DataSourceV2 by default (#64821), so readparquet and friends use the new scan/listing infrastructure with row-group-aware chunking and predicate splitting. Hash Shuffle V2 eliminates the aggregat
  • HashShuffleV2 supports join (#63598, #64538, #64687). This lets shuffles reuse standard map/reduce scheduling, backpressure, and resource accounting.
  • Ray Serve: The HAProxy ingress is now distributed as the ray-haproxy PyPI package instead of being compiled into images, and it is the default HAProxy binary (#64141, #64163, #64164). We've also added gRPC support to the HAProxy direct-ingr
  • Ray Core: We've added an embedded RocksDB storage backend for GCS fault tolerance (REP-64), selectable with RAYgcsstorage=rocksdb and RAYgcsstoragepath (#63657). GCS fault tolerance no longer requires an external Redis instance. We've also
  • 🎉 New Features
Original release notes ↗

Publicly indicated next

Prereleases and milestones indicate public plans; they are not delivery commitments.

Observation trail

  1. changedupdatedAt
  2. changedupdatedAt, metrics
  3. changedmetrics
  4. discoveredinitial snapshot
  5. discoveredinitial snapshot
  6. changedtext, updatedAt
  7. changedupdatedAt
  8. discoveredinitial snapshot
  9. discoveredinitial snapshot
  10. discoveredinitial snapshot
  11. changedupdatedAt, metrics
  12. changedupdatedAt
  13. discoveredinitial snapshot
  14. changedupdatedAt, metrics
  15. changedupdatedAt, metrics
  16. changedupdatedAt, metrics, state
  17. discoveredinitial snapshot
  18. changedupdatedAt, metrics
  19. discoveredinitial snapshot
  20. changedinitial snapshot
  21. changedinitial snapshot
  22. changedinitial snapshot
  23. changedinitial snapshot
  24. changedinitial snapshot
  25. changedinitial snapshot
  26. changedinitial snapshot
  27. changedinitial snapshot
  28. changedinitial snapshot
  29. changedinitial snapshot
  30. changedinitial snapshot
14 shown
pull request

[doc] Record what blocks lifting the setuptools ceiling and replacing sphinxcontrib-redoc

> **Stacked PR.** Fourth in a chain: #65456 (pin ReDoc) → #65457 (spec validity) → #65458 (declare 3.0.3) → this. It's independent in substance and touches only `doc/requirements-doc.txt`, so it can be retargeted to `master` if you'd rather take it on its own. ## Why this change is needed The `setuptools==80.9.0` pin in `doc/requirements-doc.txt` exists beca

importance 2@dstrodtmanclosedOriginal evidence ↗
pull request

[doc] Render the Jobs API spec with sphinxcontrib-openapi instead of ReDoc

> **Stacked PR.** Targets `doc-1513-declare-openapi-303` (#65458), which targets #65457. Review order is #65457 → #65458 → this. GitHub retargets automatically as each merges. > > **Conflicts with #65456 by design.** That PR pins `redoc_uri` to a specific ReDoc version; this one deletes `redoc_uri` entirely. If #65456 lands first I'll resolve the conflict he

importance 2@dstrodtmanopenOriginal evidence ↗
pull request

[doc] llms.txt: tell agents the page links also serve Markdown

## Summary Follow-up to #64458. Adds one line to `llms.txt` telling agents that the page links also serve Markdown. Read the Docs serves a Markdown rendering of any page **from that page's own `.html` URL**, under an `Accept: text/markdown` request. There is no separate `.md` file to link to — `page.md`, `page.html.md`, `/md/<page>.md`, and `?format=md` all

importance 2@ronny-anyscaleopengodocs-goOriginal evidence ↗
pull request

[docs] Sync the vendored KubeRay CRD API reference

## Why are these changes needed? Vendors the generated KubeRay CRD API reference at `doc/source/cluster/kubernetes/references/api.md` from `ray-project/kuberay` release tag v1.6.2, pinned at commit [`0a9eabd`](https://github.com/ray-project/kuberay/blob/0a9eabde561ab59ab4bbe626072c359da9214484/docs/reference/api.md). Opened automatically. The upstream file i

importance 2@dstrodtmanopendocsdocs-goOriginal evidence ↗
pull request

[Data] [DO NOT MERGE] arrow-rs Parquet reader A/B — treatment arm (release-test trigger)

Draft PR whose only purpose is to trigger the release pipeline for the multi-node arrow-rs vs PyArrow A/B (the reader from #65117, ported onto #64985's planner). This branch is stacked on unmerged #64985, so its diff includes that PR's commits — do not review or merge. Baseline arm: see the companion draft PR for `arrow-rs-ab-baseline-64985`. 🤖 Generated wit

importance 1@AarryaSarafopendatarelease-testOriginal evidence ↗
pull request

Bump transformers from 4.36.2 to 5.5.0 in /python

Bumps [transformers](https://github.com/huggingface/transformers) from 4.36.2 to 5.5.0. <details> <summary>Release notes</summary> <p><em>Sourced from <a href="https://github.com/huggingface/transformers/releases">transformers's releases</a>.</em></p> <blockquote> <h1>Release v5.5.0</h1> <!-- raw HTML omitted --> <h2>New Model additions</h2> <h3>Gemma4</h3>

importance 0@dependabot[bot]staledependenciespythontraincoreOriginal evidence ↗
pull request

[Data] [Core] Deferring gRPC Frees and OSS callbacks to Reduce OSM Usage Update Latency

## Description Move `free_object_on_nodes_async_` (gRPC free broadcast) and out-of-scope callbacks outside of `mutex_` in `ReferenceCounter::OnObjectOutOfScopeOrFreed`. Previously, when many `ObjectRefs` are freed simultaneously (e.g., LIMIT operators, UDF failures), the gRPC calls (~70% of `RemoveLocalReference` runtime) and callbacks (~7%) executed under `

importance 0@rayhhomeopencoredatagoOriginal evidence ↗
pull request

[llm][ci] Upgrade to vllm 0.27.0

## Description > Briefly describe what this PR accomplishes and why it's needed. ## Related issues > Link related issues: "Fixes #1234", "Closes #1234", or "Related to #1234". ## Additional information > Optional: Add implementation details, API changes, usage examples, screenshots, etc.

importance 0@jeffreywang88opengoOriginal evidence ↗
pull request

[TPU] Add subslice_index to subslice_placement_group

## Description When requesting multiple subslices simultaneously, it can be helpful to allow selecting specific subslices to guarantee a spread of jobs. This change adds an optional subslice_index parameter to the subslice_placement_group builder which restricts placement group selection. ## Related issues Follow on to #64578 ## Additional information n/a

importance 0@spencer-popentraincorecommunity-contributionOriginal evidence ↗
pull request

[doc][TPU] Document TPU subslicing options with Ray

## Description This PR documents the three methods available for TPU subslicing with Ray: 1. Dynamic slicing with Kueue (provisions subslices via GKE) 2. RayClusters with a subslice annotation 3. SubslicePlacementGroup from the Ray TPU driver. ## Related issues - [Dynamic slicing docs](https://docs.cloud.google.com/kubernetes-engine/docs/concepts/dynamic-sli

importance 0@spencer-popendocscorecommunity-contributionOriginal evidence ↗
pull request

[data] Cap task pool resource demand by runnable inputs

## Description `TaskPoolMapOperator.min_max_resource_requirements()` reports infinite maximum demand when `max_concurrency` is unset. That is a useful capacity bound, but it is not always a useful fair-share bound: an operator with one active or queued input can run at most one task. This matters for pipelines where one task fans out into many rows. The foll

importance 0@praateekmahajanopenOriginal evidence ↗
pull request

[data] Avoid rebuilding row-sized output buffers

## Description When only `target_num_rows_per_block` is set, `BlockOutputBuffer` currently builds the buffered block, emits one slice, and puts the remainder into a new builder. A StreamingRepartition with a one-row target repeats that work for every output block. This PR keeps the built block and advances through it by row offset. A builder is only needed a

importance 0@praateekmahajanopenOriginal evidence ↗
pull request

[doc] Add a KubeRay nav section and generalize the section sidebar

## Why The KubeRay guides are one of the most-visited parts of the Ray docs, but they're buried two levels down under Ray Clusters and share the global sidebar with the entire site. A reader working through RayService configuration gets every top-level Ray section in the left nav and no KubeRay-specific orientation. This promotes KubeRay to a top-level nav t

importance 0@dstrodtmanopenOriginal evidence ↗