1 Release published by 1 person

• ray-2.47.1 Ray-2.47.1

  published Jun 18, 2025

77 Pull requests merged by 38 people

• [Serve.llm][P/D] Fix health check in prefill disagg

  #53937 merged Jun 22, 2025

• [Test][KubeRay] Update KubeRay version to v1.4.0 for autoscaler tests

  #53974 merged Jun 22, 2025

• [core] Fix ActorClass.remote return typing and expose Actor class methods to static analysis

  #53986 merged Jun 21, 2025

• [core] Use core worker client pool in GCS

  #53654 merged Jun 21, 2025

• [core] Revert container tests to medium size instance

  #53966 merged Jun 21, 2025

• Fix ray import error when both ROCR_VISIBLE_DEVICES and HIP_VISIBLE_DEVICES are set

  #53757 merged Jun 20, 2025

• [core] Making NodeManager use ILocalTaskManager instead of TaskManager.

  #53961 merged Jun 20, 2025

• defer loading csat so gtag loads first

  #53968 merged Jun 20, 2025

• fix ga4 events

  #53967 merged Jun 20, 2025

• [train][template] Remove clock emoji which does not always render well

  #53965 merged Jun 20, 2025

• [core][gpu-objects] Support ray.get on the driver process for GPU objects

  #53902 merged Jun 19, 2025

• [kuberay] Update helm install command in prometheus doc to set serviceMonitor release=prometheus

  #53952 merged Jun 19, 2025

• [Docs] Fix async code in serving notebook

  #53864 merged Jun 19, 2025

• [core][rocm] Allow CUDA_VISIBLE_DEVICS and HIP_VISIBLE_DEVICES

  #53531 merged Jun 19, 2025

• [train][template] Pip install with python block instead

  #53928 merged Jun 19, 2025

• [Data] Refactor Planner to avoid storing plan-specific state

  #53955 merged Jun 19, 2025

• [core] Avoid unnecessary deserialization/serialization of CallerWorkerId

  #53939 merged Jun 19, 2025

• [serve] add ability to track child requests

  #53941 merged Jun 19, 2025

• [Doc][KubeRay] Add a doc for scheduler plugins

  #53846 merged Jun 19, 2025

• [core][telemetry/08] record counter metric e2e

  #53449 merged Jun 19, 2025

• [HashShuffle] - Add warnings for when there are insufficient resources for Aggregators

  #53705 merged Jun 19, 2025

• [Data] Join release tests

  #53903 merged Jun 19, 2025

• [docs][Serve] Add clarification for health check and FT of serve deployments

  #53944 merged Jun 19, 2025

• fastapi and streaming tests use get applications api

  #53949 merged Jun 19, 2025

• [RLlib; docs] Fix docstring example for custom MultiRLModule with shared encoder.

  #53912 merged Jun 19, 2025

• [Data] Prevent filename collisions on write

  #53890 merged Jun 19, 2025

• [data] fix flakey schema

  #53901 merged Jun 19, 2025

• [Data] Fixed BlockMetadata derivation for Read operator

  #53908 merged Jun 19, 2025

• [core] Fix flaky test_worker_exit_intended_user_exit

  #53909 merged Jun 19, 2025

• fix the bash code run error in notebook

  #53900 merged Jun 19, 2025

• [Docs] Fix issues with e2e audio tutorial

  #53932 merged Jun 19, 2025

• [train] Cleanups for training ingest benchmark

  #53684 merged Jun 19, 2025

• [train] add proper filtering to metrics

  #53788 merged Jun 18, 2025

• [cgraph] Avoid depending on torch CPU module for CPU-only actor

  #53849 merged Jun 18, 2025

• [train] expose training input/output in callbacks

  #53869 merged Jun 18, 2025

• Skip test_metrics_agent_with_open_telemetry on mac

  #53917 merged Jun 18, 2025

• [Docs] Add ServiceMonitor section and make some step optional in Grafana & Promethus page

  #53474 merged Jun 18, 2025

• [Docs][KubeRay] Update KubeRay operator installation references for all docs

  #53885 merged Jun 18, 2025

• [Core] Support AMD GPU MI3xx product line

  #51802 merged Jun 18, 2025

• [Doc][KubeRay] Update KubeRay operator installation reference

  #53842 merged Jun 18, 2025

• [Docs][KubeRay] Fix RayJob quickstart doc step 9 error

  #53887 merged Jun 18, 2025

• [Core] Use fd instead of handle for windows log redirection

  #53852 merged Jun 18, 2025

• Add dashboard visualizations for TPU metrics

  #53898 merged Jun 18, 2025

• [ObjectStore] Warn if object store is allocated < 50% of total memory for data workloads

  #53857 merged Jun 18, 2025

• [Data] Deprecate use_polars flag

  #53867 merged Jun 17, 2025

• [data] split test_all_to_all.py

  #53865 merged Jun 17, 2025

• add missing configs for object detection template

  #53895 merged Jun 17, 2025

• [core] Remove hardcoded flaky tests

  #53888 merged Jun 17, 2025

• [Serve][LLM] Simplify _prepare_engine_config()

  #53704 merged Jun 17, 2025

• [core][gpu-objects] Fix test_gpu_objects_nccl.py

  #53874 merged Jun 17, 2025

• [RLlib] MetricsLogger: Fix get/set_state to handle tensors in self.values.

  #53514 merged Jun 17, 2025

• [Data] Improve handling of mismatched columns

  #53861 merged Jun 17, 2025

• Fix pickle error with remote code models in vLLM Ray worker process

  #53815 merged Jun 17, 2025

• [train][template] Remove ineffective post build script and pip install instead

  #53822 merged Jun 17, 2025

• [core][gpu objects] Integrate single-controller collective APIs with GPU objects

  #53720 merged Jun 16, 2025

• [Data] Improve handling of pandas.NA

  #53859 merged Jun 16, 2025

• [devx] Fix 'uv run' command line parsing

  #53838 merged Jun 16, 2025

• [Data] Improve read_text trailing newline semantics

  #53860 merged Jun 16, 2025

• [Serve.llm][P/D] Support separate deployment config for PDProxy in Prefill disagg

  #53821 merged Jun 16, 2025

• [Doc][KubeRay] Remove vllm-rayservice.md and use Ray Serve LLM instead

  #53844 merged Jun 16, 2025

• add api to get application url

  #53796 merged Jun 16, 2025

• [Doc][KubeRay] Remove very old ResNet benchmark example

  #53839 merged Jun 16, 2025

• [release] Fix release tests

  #53855 merged Jun 16, 2025

• [Serve.llm] Disable TP=2 VLM batch test

  #53825 merged Jun 16, 2025

• [Doc][Fix] reveal the falsely hidden export command in the KubeRay GCS FT guide

  #53832 merged Jun 16, 2025

• [core][gpu-objects] Support intra-process communication

  #53798 merged Jun 16, 2025

• [Doc][KubeRay] Remove very old XGBoostTrainer example

  #53837 merged Jun 16, 2025

• [core] Release resources only after tasks have stopped executing

  #53660 merged Jun 16, 2025

• [core] Deflake test_multiprocessing.py

  #53802 merged Jun 16, 2025

• [core] Fix test_object_spilling.py on Windows

  #53851 merged Jun 16, 2025

• [KubeRay] Remove unused YAMLs

  #53840 merged Jun 16, 2025

• [chore] Change file mode of rayservice-no-ray-serve-replica.md from 755 to 644

  #53843 merged Jun 16, 2025

• fix AggregateFnV2 doc to state finalize instead of _finalize

  #53835 merged Jun 16, 2025

• [core] Fix GCS subscribers map race condition

  #53781 merged Jun 16, 2025

• [core] deleting unused code from plasma client

  #53814 merged Jun 16, 2025

• [core] Fix race condition in raylet graceful shutdown

  #53762 merged Jun 16, 2025

• [serve] Revert request timeout from serve instance fixtures

  #53809 merged Jun 16, 2025

61 Pull requests opened by 37 people

• [core] Ungracefully exit if the agent dies unexpectedly

  #53847 opened Jun 16, 2025

• [RLlib] Mixin Layer Design Sketch Up

  #53850 opened Jun 16, 2025

• [core] Fix comment

  #53853 opened Jun 16, 2025

• [Docs] Add image tag to `rayproject/ray-ml`

  #53854 opened Jun 16, 2025

• [core] adding additional stats to the dump object store usage api.

  #53856 opened Jun 16, 2025

• [core] Cleanup naming in core worker scheduling queues

  #53858 opened Jun 16, 2025

• [core] Sleep to debug container test

  #53862 opened Jun 16, 2025

• [core] Don't queue in flight submissions by attempt number

  #53866 opened Jun 16, 2025

• Feat/ray serve middleware support

  #53868 opened Jun 17, 2025

• Pass parameters to custom routers through LLMConfig

  #53870 opened Jun 17, 2025

• test: refactor `test_observability_helpers`

  #53875 opened Jun 17, 2025

• [dashboard] Support to overwrite the _client_max_size of http request entity

  #53880 opened Jun 17, 2025

• [doc][core] fix reStructuredText formatting on Resources page

  #53882 opened Jun 17, 2025

• [Docs][KubeRay] Update all KubeRay version references for KubeRay 1.4.0 release

  #53884 opened Jun 17, 2025

• [Docs][KubeRay] Update changes from KubeRay 1.3.2 to 1.4.0

  #53886 opened Jun 17, 2025

• [ci] add python 3.13 ray docker image build

  #53894 opened Jun 17, 2025

• Bump gradio from 3.50.2 to 5.31.0 in /python/requirements

  #53899 opened Jun 17, 2025

• python depsets tool

  #53904 opened Jun 18, 2025

• [core] Move inner_publisher logic into gcsPublisher

  #53905 opened Jun 18, 2025

• [Core] Remove Unnecessary Checks in GRPC Server Shutdown Process

  #53910 opened Jun 18, 2025

• [WIP][core][gpu-objects] GC

  #53911 opened Jun 18, 2025

• [RLlib] Add missing colon to CUBLAS_WORKSPACE_CONFIG

  #53913 opened Jun 18, 2025

• [kuberay] log actionable err msg when required TPU node selectors missing

  #53914 opened Jun 18, 2025

• [RLlib] Add missing documentation for SACConfig's training()

  #53918 opened Jun 18, 2025

• [core][telemetry/12] record histogram metric e2e

  #53927 opened Jun 18, 2025

• Update deletion policy for rayjob quick start

  #53929 opened Jun 18, 2025

• [Data] - write_parquet enable both partition by & min_rows_per_file, max_rows_per_file

  #53930 opened Jun 18, 2025

• [core][telemetry/13] performance tests

  #53931 opened Jun 18, 2025

• [serve] move test from test_grpc to test_proxy

  #53933 opened Jun 18, 2025

• [core] Fix race condition b/w object eviction & repinning for recovery.

  #53934 opened Jun 18, 2025

• [core][GPU objects] Attach tensor transport to task args protobuf

  #53935 opened Jun 18, 2025

• Bump urllib3 from 1.26.19 to 2.5.0 in /python

  #53936 opened Jun 18, 2025

• [V2][Autoscaler] Transition ALLOCATED instances to RAY_STOPPPED when enforcing max nodes

  #53938 opened Jun 18, 2025

• [Data] Replaced `get_object_locations` with `get_local_object_locations`

  #53942 opened Jun 19, 2025

• [doc][kuberay] state `rayStartParams` is optional starting with KubeRay 1.4.0

  #53943 opened Jun 19, 2025

• [doc][kuberay] add version skew warning for plugin and RayCluster

  #53950 opened Jun 19, 2025

• [core] Fix GCS crash on duplicate MarkJobFinished RPCs due to network failures

  #53951 opened Jun 19, 2025

• [data] remove schema from release tests

  #53956 opened Jun 19, 2025

• [serve.llm] Prefix aware router eviction thread improvements

  #53957 opened Jun 19, 2025

• [core] improve assertion check in test_task_metrics

  #53958 opened Jun 19, 2025

• Can wins01

  #53959 opened Jun 19, 2025

• [core] disable so_reuseaddr when creating grpc server

  #53960 opened Jun 19, 2025

• [data] remove operator_fusion_benchmark

  #53962 opened Jun 19, 2025

• [train] Fix release test missing data key

  #53963 opened Jun 19, 2025

• finishing commit for issue #52113

  #53964 opened Jun 19, 2025

• tune: make Tune status/progress tables readable in dark mode

  #53969 opened Jun 20, 2025

• [Data] Fixing PA overflow handling

  #53971 opened Jun 20, 2025

• docs(data): fix broken Parameters table

  #53972 opened Jun 20, 2025

• [core] Rename `GcsFunctionManager` and use fake in test

  #53973 opened Jun 20, 2025

• [core] Fix flaky `test_state_api`

  #53975 opened Jun 20, 2025

• Revert "[train] Cleanups for training ingest benchmark"

  #53979 opened Jun 20, 2025

• [Serve.llm] Remove ImageRetriever class and related tests from the LLM deployment module.

  #53980 opened Jun 20, 2025

• Feature/sac discrete

  #53982 opened Jun 20, 2025

• [Data] Fix ActorPool autoscaler to properly scale up

  #53983 opened Jun 20, 2025

• [CI][KubeRay] Update KubeRay CI Tests branch for KubeRay v1.4.0 release

  #53984 opened Jun 21, 2025

• [Core] Add AcceleratorManager implementation for Rebellions NPU

  #53985 opened Jun 21, 2025

• [Doc] Update Istio service mesh graph

  #53988 opened Jun 21, 2025

• [Serve] Make replica scheduler backoff configurable #52871

  #53991 opened Jun 21, 2025

• Fix autoscaler recovery docker config to use node-specific settings

  #53992 opened Jun 21, 2025

• [ci] Upgrade nightly test to run against KubeRay 1.4

  #53993 opened Jun 21, 2025

• [core] Remove actor task path in normal task submitter

  #53996 opened Jun 22, 2025

46 Issues closed by 23 people

• CI test linux://python/ray/data:test_arrow_block is flaky

  #48859 closed Jun 22, 2025

• Conflict between ROCR_VISIBLE_DEVICES and HIP_VISIBLE_DEVICES environment variables causes Ray import error

  #53737 closed Jun 21, 2025

• CI test linux://rllib:learning_tests_cartpole_dqn_multi_cpu is flaky

  #47214 closed Jun 21, 2025

• CI test linux://rllib:examples/evaluation/evaluation_parallel_to_training_multi_agent_duration_auto is consistently_failing

  #53255 closed Jun 21, 2025

• CI test linux://python/ray/serve/tests:test_multiplex is flaky

  #48378 closed Jun 21, 2025

• CI test windows://python/ray/tests:test_object_store_metrics is flaky

  #49514 closed Jun 20, 2025

• [RLlib] MAML does not work with TF2 in Ray 2.3.1

  #34620 closed Jun 20, 2025

• [RayData|RayServe] Does RayData/RayServe support multi-node vllm inference

  #53192 closed Jun 20, 2025

• [Core] Core Worker crashing

  #49088 closed Jun 20, 2025

• [core][gpu-objects] Driver tries to get the data from in-actor store

  #51272 closed Jun 19, 2025

• [Core][ROCm] Setting CUDA_VISIBLE_DEVICES leads to an assertion

  #52701 closed Jun 19, 2025

• [Autoscaler][V2] Autoscaler fails to delete idle KubeRay Pod

  #52264 closed Jun 19, 2025

• CI test linux://python/ray/data:test_consumption is flaky

  #48163 closed Jun 19, 2025

• CI test windows://python/ray/tests:test_actor_state_metrics is consistently_failing

  #46303 closed Jun 19, 2025

• [data] ray.data.read_images is slower than reading images manually

  #37499 closed Jun 19, 2025

• [RFC] Q2 Ray Data Roadmap

  #51808 closed Jun 19, 2025

• [RFC] LLM APIs for Ray Data and Ray Serve

  #50639 closed Jun 19, 2025

• CI test linux://python/ray/data:test_json is flaky

  #48150 closed Jun 19, 2025

• [<Ray component: Core|RLlib|etc...>] ValueError: There was an error while reducing the Stats object under key=('actual_n_step',)!

  #53947 closed Jun 19, 2025

• CI test windows://python/ray/serve/tests:test_standalone_3 is flaky

  #44003 closed Jun 19, 2025

• Release test compiled_graphs failed

  #53716 closed Jun 18, 2025

• CI test darwin://python/ray/tests:test_metrics_agent_open_telemetry is consistently_failing

  #53828 closed Jun 18, 2025

• [RLlib] ActionMaskingTorchRLModule can't set up `conv_filters`

  #53325 closed Jun 18, 2025

• [Core] `ray.init()` and `ray start` fails on Windows 11 in ray 2.45+

  #52739 closed Jun 18, 2025

• CI test windows://python/ray/tests:test_object_spilling_debug_mode is flaky

  #43796 closed Jun 18, 2025

• [core] support S3 path style access in runtime_env download_and_unpack_package()

  #53893 closed Jun 17, 2025

• How to transfer tensors stored in GPU in actor with NCCL?

  #53816 closed Jun 17, 2025

• [Data] PyArrow 20.0.0 Backward Incompatability (`unexpected keyword argument 'maps_as_pydicts'`)

  #52685 closed Jun 17, 2025

• CI test linux://python/ray/tests:test_gpu_objects_nccl is consistently_failing

  #53871 closed Jun 17, 2025

• [RLlib] Headnode without GPU triggers torch/CUDA de-serialization error

  #53467 closed Jun 17, 2025

• [Core] Ray Autoscaler does not restart a worker node on setup failure

  #29127 closed Jun 17, 2025

• Release test llm_batch_vllm failed

  #53827 closed Jun 17, 2025

• [Serve] Add timeout parameter for `deploy`

  #25433 closed Jun 17, 2025

• CI test windows://python/ray/serve/tests:test_logging is flaky

  #46043 closed Jun 17, 2025

• [Core] Read-only buffer error in some scikit-learn models

  #52571 closed Jun 17, 2025

• [core] ray stop --force doesn't kill processes on worker node

  #28038 closed Jun 17, 2025

• [core][gpu-objects] Support TensorDict

  #51550 closed Jun 17, 2025

• [core][gpu-objects] Allocate placeholder tensor on corresponding devices

  #53622 closed Jun 17, 2025

• [core][gpu-objects] Driver should order all collective calls to avoid deadlock

  #51264 closed Jun 17, 2025

• CI test windows://python/ray/tests:test_object_spilling_asan is consistently_failing

  #45962 closed Jun 17, 2025

• CI test windows://python/ray/tests:test_object_spilling is consistently_failing

  #45961 closed Jun 16, 2025

• [RLlib] Add syntax checking to configuration string literals or migrate to enums.

  #39384 closed Jun 16, 2025

• [Ray Core] Ray error causes the Python interpreter to terminate without failing

  #28211 closed Jun 16, 2025

• [CI] Test GPU training tutorial with Ray Release tests

  #28902 closed Jun 16, 2025

• CI test linux://python/ray/train:accelerate_torch_trainer is consistently_failing

  #44513 closed Jun 16, 2025

• [core][gpu-objects] intra-process communication

  #51685 closed Jun 16, 2025

36 Issues opened by 31 people

• [Data] When writing on BigQuery, Google's "TooManyRequests" exceptions is not retried

  #53997 opened Jun 22, 2025

• [data] Slow fetching of metadata for large number of parquet files

  #53995 opened Jun 22, 2025

• [Rllib] Bug in TorchMultiDistribution logp prevents policy mapping from being used

  #53994 opened Jun 22, 2025

• Release test many_nodes_actor_test_on_v2.aws failed

  #53990 opened Jun 21, 2025

• [Doc][KubeRay] Run doctest `user-guides/configuring-autoscaling.ipynb` in CI

  #53989 opened Jun 21, 2025

• [Core] Autoscaler Node Recovery Ignores Node-Specific Docker Config

  #53987 opened Jun 21, 2025

• [core][gpu-objects] Allow sending ObjectRefs to other processes

  #53978 opened Jun 20, 2025

• [core][gpu-objects] Support ray.put

  #53977 opened Jun 20, 2025

• [core][gpu-objects] RDMA support for data transfer

  #53976 opened Jun 20, 2025

• [Dashboard] Support for List Tasks Filter Pushdown

  #53970 opened Jun 20, 2025

• [Data] Add support to turn off strict block-size enforcement

  #53954 opened Jun 19, 2025

• Release test training_ingest_benchmark-task=image_classification.full_training.jpeg failed

  #53953 opened Jun 19, 2025

• ValueError: There was an error while reducing the Stats object under key=('actual_n_step',)! [<Ray component: Core|RLlib|etc...>]

  #53948 opened Jun 19, 2025

• [Core] `InternalKVPut` retries incorrectly when encountering transient error

  #53946 opened Jun 19, 2025

• PolicyServer and PolicyClient Demo Issue

  #53926 opened Jun 18, 2025

• CI test linux://rllib:examples/algorithms/vpg_custom_algorithm is flaky

  #53925 opened Jun 18, 2025

• Windows VS WSL2

  #53924 opened Jun 18, 2025

• [Docker][CI] Add Python 3.13 Ray Image to CI

  #53923 opened Jun 18, 2025

• [serve.llm] Ray LLM serving not respecting max_completion_tokens parameter

  #53922 opened Jun 18, 2025

• [Ray V2 Tune + Train] Tuner is not aware of resources and oversubscribes leading to deadlocks

  #53921 opened Jun 18, 2025

• [Data/Preprocessors]: Preprocessors do not work with nested records

  #53920 opened Jun 18, 2025

• [Core] Ray Does Not Detect GPU

  #53919 opened Jun 18, 2025

• Multiple CVEs in Ray's compiled dependencies

  #53915 opened Jun 18, 2025

• Using ray for LLM inference got errors

  #53907 opened Jun 18, 2025

• [Core] Starting multiple local instances on one node may result in errors due to randomly selecting the same port.

  #53906 opened Jun 18, 2025

• [CI] `linux://python/ray/data:test_consumption` is failing/flaky on master.

  #53897 opened Jun 17, 2025

• [Ray Data]Pylint detection found some Python code defects in ray data

  #53881 opened Jun 17, 2025

• [dashboard] Support to overwrite the _client_max_size of http request entity

  #53879 opened Jun 17, 2025

• [RLlib] Significant drop in DQN training reward when resuming from checkpoint

  #53878 opened Jun 17, 2025

• [RLlib] Checkpoint metrics loading with Tune is broken in 2.47.0

  #53877 opened Jun 17, 2025

• Issue: Ray Dashboard Links to Grafana Return "Dashboard Not Found" (Windows)

  #53876 opened Jun 17, 2025

• [serve.llm] LLM serving seems not working with mistral tokenizer.

  #53873 opened Jun 17, 2025

• [Core] ray.ActorID.nil().job_id

  #53872 opened Jun 17, 2025

• [Core] Ray 2.47 regression: All tasks hang when using `uv`

  #53848 opened Jun 16, 2025

• [<Ray component: Core|RLlib|etc...>] Ray Timeout Error running VLLM Multi-Node(tp_size=2) Online Server with Acl_Graph when handling curl request

  #53845 opened Jun 16, 2025

• [RLlib] Typo in error message on line 37 of ray/rllib/env/utils/__init__.py

  #53841 opened Jun 16, 2025

1,427 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• [core][compiled graphs] Supporting allreduce on list of input nodes

  #51047 commented on Jun 20, 2025 • 19 new comments

• [Doc][KubeRay] Add doc for running KubeRay dashboard

  #53830 commented on Jun 21, 2025 • 17 new comments

• [Feat][Core] Implement Event Aggregator Agent

  #53182 commented on Jun 19, 2025 • 14 new comments

• [train] Driver SIGINT calls controller abort

  #53600 commented on Jun 19, 2025 • 13 new comments

• [train] TrainStateActor periodically checks controller status and sets aborted

  #53818 commented on Jun 20, 2025 • 12 new comments

• Add `pin_memory` to `iter_torch_batches`

  #53792 commented on Jun 20, 2025 • 10 new comments

• [core] fix detached actor being unexpectedly killed

  #53562 commented on Jun 20, 2025 • 10 new comments

• (serve.llm) Make _LLMServerBase.__init__ synchronous

  #53719 commented on Jun 16, 2025 • 3 new comments

• (serve.llm): Refactor/Consolidate LoRA downloading

  #53714 commented on Jun 19, 2025 • 3 new comments

• Add Apple silicon GPU(mps) support to ray

  #38464 commented on Jun 20, 2025 • 2 new comments

• Relax check_version_info to check for bytecode compatibility

  #41373 commented on Jun 19, 2025 • 2 new comments

• [Core] Add default Ray Node labels at Node init

  #53360 commented on Jun 21, 2025 • 2 new comments

• [core] Support broadcast and reduce collective for compiled graphs

  #53625 commented on Jun 18, 2025 • 2 new comments

• ray: fix handling large chunks

  #53535 commented on Jun 20, 2025 • 1 new comment

• [core]: Correct podman output parsing for image uri in runtime env

  #53653 commented on Jun 18, 2025 • 1 new comment

• [Core] The Idle worker killing feature slows down tasks

  #27863 commented on Jun 17, 2025 • 0 new comments

• [RLlib] Issue Regarding Future Warnings

  #26424 commented on Jun 16, 2025 • 0 new comments

• [Ray component: Core] Enable better progress bar

  #26426 commented on Jun 16, 2025 • 0 new comments

• [core][c++ worker] RayClusterModeTest.DefaultActorLifetimeTest timed out in macOS

  #26435 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Use observations (input_dict) for exploration

  #26437 commented on Jun 16, 2025 • 0 new comments

• [Core] Observing Multiple Exceptions When Using Different Python Patch Versions

  #26443 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Unable to call ray.remote functions inside env/action dist

  #26468 commented on Jun 16, 2025 • 0 new comments

• [Job] Job submission not following convention for quote

  #26514 commented on Jun 16, 2025 • 0 new comments

• [doc][Core | State Observability] Document usage of the rate limiting env variable in public doc

  #26370 commented on Jun 16, 2025 • 0 new comments

• [Tune] NevergradSearch Budget Exception

  #26305 commented on Jun 16, 2025 • 0 new comments

• How to color-code console output

  #26226 commented on Jun 16, 2025 • 0 new comments

• [Core] [Quality] Live handle raises unnecessary exception when script ends

  #26198 commented on Jun 16, 2025 • 0 new comments

• [RLlib]: SimpleQ TF2 is broken

  #26192 commented on Jun 16, 2025 • 0 new comments

• [State Observability] Raise an exception if the state schema contains predicates.

  #26125 commented on Jun 16, 2025 • 0 new comments

• [air] We should have a convenient method for user to interact with checkpoint file on driver when they checkpoint using other method in session

  #26082 commented on Jun 16, 2025 • 0 new comments

• [RLlib] server reports nan episodes and empty policy

  #26048 commented on Jun 16, 2025 • 0 new comments

• [test][autoscaler] ModuleNotFoundError: No module named 'ray.tests'

  #26023 commented on Jun 16, 2025 • 0 new comments

• [Tune] Ray Tune doesn't work inside Spark UDF

  #26002 commented on Jun 16, 2025 • 0 new comments

• Ray component: Core: PoolActor processes hanging

  #24784 commented on Jun 16, 2025 • 0 new comments

• [AIR] Support TorchRec trainer

  #27794 commented on Jun 17, 2025 • 0 new comments

• [Dashboard] Dashboard agent cannot be started because the port is still occupied

  #27736 commented on Jun 17, 2025 • 0 new comments

• [Tune/RLlib] log_to_file creates files, but doesn't write anything there

  #27702 commented on Jun 17, 2025 • 0 new comments

• [RLLib] global_timestep not monotonic when when running concurrent episodes with ExternalEnv

  #27669 commented on Jun 17, 2025 • 0 new comments

• [Dashboard] Ray Dashboard not showing the SpillWorker's actual memory usage

  #27591 commented on Jun 17, 2025 • 0 new comments

• [Core] The actors got distributed to just a few nodes even with spread scheduling

  #27577 commented on Jun 17, 2025 • 0 new comments

• [runtime_env] Add tests for all driver output (warnings, etc)

  #27566 commented on Jun 17, 2025 • 0 new comments

• [Tune] TuneReportCheckpointCallback causes two checkpoints to made every time it is called.

  #27524 commented on Jun 17, 2025 • 0 new comments

• [AIR] SettingWithCopyWarning for "A value is trying to be set on a copy of a slice from a DataFrame"

  #27352 commented on Jun 17, 2025 • 0 new comments

• [ray dashboard] profile button not working

  #27211 commented on Jun 17, 2025 • 0 new comments

• [Ray Train] Ray Train running slow when multiple workers executed

  #27107 commented on Jun 17, 2025 • 0 new comments

• [workflow] We should give the storage a default value if it's not set in some way.

  #27046 commented on Jun 17, 2025 • 0 new comments

• [State Observability][Log] Allow to ctrl + C when running logs API

  #27008 commented on Jun 17, 2025 • 0 new comments

• [runtime env] local `working_dir` doesn't work with strongly-typed `RuntimeEnv`

  #26984 commented on Jun 17, 2025 • 0 new comments

• [Core][State Observability] More fine-grained exceptions/error codes handling

  #26974 commented on Jun 17, 2025 • 0 new comments

• [Feature] Autoscaler should understand AWS availability and act accordingly

  #20774 commented on Jun 16, 2025 • 0 new comments

• [Core] Typing for .options for Ray Tasks

  #26871 commented on Jun 16, 2025 • 0 new comments

• [State Observability] Support filter None value

  #26820 commented on Jun 16, 2025 • 0 new comments

• [Core] Batch PinObjectIDs requests from Raylet client

  #26796 commented on Jun 16, 2025 • 0 new comments

• [Train] feature request for catboost_ray

  #26687 commented on Jun 16, 2025 • 0 new comments

• [AIR/Tune] Add a `ScalingConfig`-based API to `ResourceChangingScheduler`

  #26538 commented on Jun 16, 2025 • 0 new comments

• [RLlib] CRR and CQL consume more cpus than reported

  #26533 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Duplicate custom metrics

  #24731 commented on Jun 16, 2025 • 0 new comments

• [Serve] Asynchronous inference best practices

  #24627 commented on Jun 16, 2025 • 0 new comments

• [tune] `progress_reporter.py` is messy and should be cleaned up

  #24604 commented on Jun 16, 2025 • 0 new comments

• [aws][autoscaler] AWS: When using spot instances, always single availability zone is selected

  #24310 commented on Jun 16, 2025 • 0 new comments

• [RLlib] PPO - ray.rllib.agents.ppo "Put Error"

  #24307 commented on Jun 16, 2025 • 0 new comments

• [Ray Collective] Remove Redis store and LocalFile store from gloo mode.

  #24288 commented on Jun 16, 2025 • 0 new comments

• [Autoscaler] upscaling_speed: 0 gets reset to 1

  #24177 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Categorical action dist incorrectly uses tf.random.categorical

  #24055 commented on Jun 16, 2025 • 0 new comments

• [RLlib] [Bug] Inconsistent behavior between TFPolicy and TorchPolicy on `compute_actions_from_input_dict`

  #24007 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Enable Training from Replay Buffer Larger than Memory

  #23816 commented on Jun 16, 2025 • 0 new comments

• [RLlib] [Bug] IMPALA causes an OOM after a long running.

  #23769 commented on Jun 16, 2025 • 0 new comments

• [BUG] Ray dashboard client failed to build

  #23548 commented on Jun 16, 2025 • 0 new comments

• [RFC][Feature][Autoscaler][Core]Graceful draining of nodes while scale-down

  #23522 commented on Jun 16, 2025 • 0 new comments

• [ml][Improvement] Improve messages to be “rank0, rank1” actors etc.

  #23310 commented on Jun 16, 2025 • 0 new comments

• [Feature] [tune] create a mlflow run name from config params

  #23228 commented on Jun 16, 2025 • 0 new comments

• [Feature][RLlib] Improve pytorch memory usage by disabling caching

  #23077 commented on Jun 16, 2025 • 0 new comments

• [tune][Bug] Worker doesn't sync the logs to HDFS at the given interval

  #23055 commented on Jun 16, 2025 • 0 new comments

• [Bug] AdaBelief optimizer crashes checkpoint restore

  #22976 commented on Jun 16, 2025 • 0 new comments

• [Bug] Custom model with R2D2

  #22747 commented on Jun 16, 2025 • 0 new comments

• [Bug] Resources displayed in Dashboard don't match cluster configuration

  #22548 commented on Jun 16, 2025 • 0 new comments

• [Bug] Deletion of Ray clusters hangs while Ray operator is still up

  #22505 commented on Jun 16, 2025 • 0 new comments

• Doing import ray breaks my logging [Bug]

  #22312 commented on Jun 16, 2025 • 0 new comments

• [Serve] A Deployment Graph with unfulfilled demands fails to scale Pods in Kubernetes

  #25998 commented on Jun 16, 2025 • 0 new comments

• API server internal error message not useful

  #25986 commented on Jun 16, 2025 • 0 new comments

• [runtime env] Bad runtime env specified in ray.init() with eager install only raises error on task/actor invocation

  #25972 commented on Jun 16, 2025 • 0 new comments

• [Core][State Observability] Use a separate thread to run spill/restore

  #25960 commented on Jun 16, 2025 • 0 new comments

• [RLlib] KeyError: simple_list_collector.py, line 950, in postprocess_episode

  #25938 commented on Jun 16, 2025 • 0 new comments

• [RLLib] SampleBatch.update() doesn't update `added_keys`

  #25937 commented on Jun 16, 2025 • 0 new comments

• [runtime env] Use namespace for internal KV storage

  #25897 commented on Jun 16, 2025 • 0 new comments

• [Core?] Federation + data perimeters

  #25846 commented on Jun 16, 2025 • 0 new comments

• [Core] RBAC + auditability

  #25845 commented on Jun 16, 2025 • 0 new comments

• [Core] Arrow Flight Server doesn't work with Ray Actors due to two GRPC versions

  #25774 commented on Jun 16, 2025 • 0 new comments

• [Core | State Observability ] Refactor summary/log SDK to use StateApiClient

  #25746 commented on Jun 16, 2025 • 0 new comments

• [Serve] Deployment fails if name contains slashes

  #25714 commented on Jun 16, 2025 • 0 new comments

• [RLlib] ModelCatagolg Selects Wrong Model for Nested Complex Observations

  #25619 commented on Jun 16, 2025 • 0 new comments

• [Train] [Tune] When using Train with Tune, a `logdir` is created that's not the one specified by the user

  #25474 commented on Jun 16, 2025 • 0 new comments

• [Core][Observability] Ray memory should show more objects

  #25463 commented on Jun 16, 2025 • 0 new comments

• [Dashboard] Error during render node with gpu and 4 hdds

  #25437 commented on Jun 16, 2025 • 0 new comments

• [Ray Collective Lib] Enable CI

  #25396 commented on Jun 16, 2025 • 0 new comments

• Core: deamonset feature request

  #25334 commented on Jun 16, 2025 • 0 new comments

• [DeviceMesh][Collective] Support multiple tensors API

  #25129 commented on Jun 16, 2025 • 0 new comments

• [Ray Air] nan in the tensorflow_linear_dataset_example.py

  #25037 commented on Jun 16, 2025 • 0 new comments

• ray docker images do not have uvloop installed

  #25023 commented on Jun 16, 2025 • 0 new comments

• Ray Tune: No console output is logged to Wandb.

  #25011 commented on Jun 16, 2025 • 0 new comments

• [Core][RLlib][Tune] CUDA PTX error when training with Tune

  #25001 commented on Jun 16, 2025 • 0 new comments

• [Build][Deps] Add new `ray[azure]` extra package

  #48847 commented on Jun 22, 2025 • 0 new comments

• [AIR][Tune] Provide user guide on how to build active learning on AIR

  #30157 commented on Jun 17, 2025 • 0 new comments

• [AIR/Docs] Mention/warn that running a Trainer inside a custom Tune trainable is an anti-pattern

  #30153 commented on Jun 17, 2025 • 0 new comments

• [RLlib] (PPO) algo parameter "lambda_" never gets passed because `AlgorithmConfig` refractored "lambda_" to "lambda"

  #30143 commented on Jun 17, 2025 • 0 new comments

• [Core] Reference leakage somewhere after ray.shutdown()

  #30089 commented on Jun 17, 2025 • 0 new comments

• [Tune] Can't access all metrics for all trials

  #30004 commented on Jun 17, 2025 • 0 new comments

• [core][dashboard] state api on worker nodes can not connect to dashboard url

  #29959 commented on Jun 17, 2025 • 0 new comments

• [Jobs] Include requested and available resources in JobInfo status message

  #29921 commented on Jun 17, 2025 • 0 new comments

• [RLlib] Add some metric for aync algos (e.g. APPO) that shows the total number of gradient updates

  #29830 commented on Jun 17, 2025 • 0 new comments

• [AIR] Update pytorch training and prediction benchmark with numpy with updated metrics

  #29743 commented on Jun 17, 2025 • 0 new comments

• [RLlib] Undesired memory growing when using convolutional neural network

  #29699 commented on Jun 17, 2025 • 0 new comments

• [AIR] `XGBoostTrainer` gives misleading error if column missing

  #29695 commented on Jun 17, 2025 • 0 new comments

• [RLLib Tests] : Included pytests in package as well as basic commands fail with ValueError

  #29691 commented on Jun 17, 2025 • 0 new comments

• [air] GPU memory leak when using AIR trainer with torch dataloader when the latter uses multi-processing

  #29563 commented on Jun 17, 2025 • 0 new comments

• [RLlib] Benchmark bandit methods vs plain Thompson Sampling for a non-contextual MAB

  #29528 commented on Jun 17, 2025 • 0 new comments

• [RLlib] "model": {"free_log_std": True} generates Tensorflow Lambda layers warning with TF2 framework

  #29502 commented on Jun 17, 2025 • 0 new comments

• [Autoscaler] Delete AWS resources created when launching Ray cluster upon cluster termination

  #29499 commented on Jun 17, 2025 • 0 new comments

• [Core] util.multiprocessing.pool: imap and imap_unordered blocking on ray.wait even though processes are complete

  #29466 commented on Jun 17, 2025 • 0 new comments

• [Ray Log_monitor]: close_all_files ProcessLookupError

  #29452 commented on Jun 17, 2025 • 0 new comments

• Ray core: incorrect account of GPUs on ec2 ubuntu instance: g4dn.2xlarge

  #29420 commented on Jun 17, 2025 • 0 new comments

• [core] GCS segfaults under OOM

  #29336 commented on Jun 17, 2025 • 0 new comments

• [AIR] Add progress bar for training

  #29314 commented on Jun 17, 2025 • 0 new comments

• [CI] A simple way to reproduce osx/linux/windows CI run failure locally

  #29068 commented on Jun 17, 2025 • 0 new comments

• [RLlib] Deprecate the RLlib spaces that are duplications of gym spaces.

  #30800 commented on Jun 17, 2025 • 0 new comments

• [Tune] Guard against users overriding internal `Trainable` methods

  #30795 commented on Jun 17, 2025 • 0 new comments

• Ray Cluster Resources Issue

  #30780 commented on Jun 17, 2025 • 0 new comments

• [Core] Worker leak

  #30731 commented on Jun 17, 2025 • 0 new comments

• [RLlib] Default policy error in two trainer work flow

  #30676 commented on Jun 17, 2025 • 0 new comments

• [core] Can't set working directory for runtime env in actor definition

  #30666 commented on Jun 17, 2025 • 0 new comments

• [Tune] HeboSearch reproducible deterministic results

  #30661 commented on Jun 17, 2025 • 0 new comments

• [core] Memory changes are not as expected when using ray.get()

  #30615 commented on Jun 17, 2025 • 0 new comments

• [Tune] `fail_fast` marks all runs as terminated, making the experiment impossible to restore

  #30584 commented on Jun 17, 2025 • 0 new comments

• [RLLib] Custom model with LSTM causes the auto wrapping to be partially executed

  #30581 commented on Jun 17, 2025 • 0 new comments

• [Core|RayTrain] RuntimeError: Some workers returned results while others didn't

  #30545 commented on Jun 17, 2025 • 0 new comments

• [Core] Overriding the default logging format for Worker logs

  #30544 commented on Jun 17, 2025 • 0 new comments

• [AIR] Canonical way to determine whether the code is running in a Train/Tune session

  #30536 commented on Jun 17, 2025 • 0 new comments

• [client][runtime_env] Inconsistent runs on ray client

  #30518 commented on Jun 17, 2025 • 0 new comments

• [Core] ray.exceptions.RaySystemError: System error: buffer source array is read-only

  #30505 commented on Jun 17, 2025 • 0 new comments

• [Core] Access violation on windows 11 when running modin workload

  #30493 commented on Jun 17, 2025 • 0 new comments

• Critic Regularized Regression (CRR) model is getting error with Custom Environment (Offline RL)

  #30411 commented on Jun 17, 2025 • 0 new comments

• [Docs] [Jobs] Add pros and cons of different ways of submitting a job

  #30305 commented on Jun 17, 2025 • 0 new comments

• [air/horovod] horovod distributed worker creation may hang

  #30276 commented on Jun 17, 2025 • 0 new comments

• [<Ray component: Workflow>] module 'ray.workflow' has no attribute 'HTTPListener'

  #30248 commented on Jun 17, 2025 • 0 new comments

• [RLLIB][Torch] numerically unstable + mkl issue in torch.sqrt normc_initializer

  #30191 commented on Jun 17, 2025 • 0 new comments

• [RLlib] RuntimeError: invalid multinomial distribution (sum of probabilities <= 0)

  #30164 commented on Jun 17, 2025 • 0 new comments

• [Core] [RLlib] RLlib on Ray 2.0 not easily working on Colab

  #28457 commented on Jun 17, 2025 • 0 new comments

• [Job Submission] Support env file input in ray.runtime_env.RuntimeEnv

  #28453 commented on Jun 17, 2025 • 0 new comments

• [Tune] Adding DEHB

  #28427 commented on Jun 17, 2025 • 0 new comments

• [serve] Gradio integration does surface error messages, runs indefinitely

  #28399 commented on Jun 17, 2025 • 0 new comments

• [Ray: Core] Ray can hang when getting an ObjectRef from an unknown environment

  #28341 commented on Jun 17, 2025 • 0 new comments

• [Core][RuntimeEnv]Make `job_submission_id` to a new index of GCS::JobTableData

  #28337 commented on Jun 17, 2025 • 0 new comments

• [Jobs] Run jobs tests on Windows

  #28316 commented on Jun 17, 2025 • 0 new comments

• [Core] Job stop should terminate runtime_env setup

  #28221 commented on Jun 17, 2025 • 0 new comments

• [Core] log_to_driver=False does not suppress worker errors in ipython

  #28216 commented on Jun 17, 2025 • 0 new comments

• [core][runtime envs] Ray should respect CUDA_VISIBLE_DEVICES if set in runtime env

  #28215 commented on Jun 17, 2025 • 0 new comments

• [Core] ray dashboard <rayhost>:8265/nodes?view=details cpuPercent should contains actor's subprocess

  #28100 commented on Jun 17, 2025 • 0 new comments

• [Core] Ray may hang if workers fail to start due to limited ports

  #28071 commented on Jun 17, 2025 • 0 new comments

• [<Ray component: Core|Cluster>] Documentation instructions for mounting AWS EFS Fails for Ray Cluster

  #28057 commented on Jun 17, 2025 • 0 new comments

• [Core] Support retry_delay option in Ray tasks

  #28015 commented on Jun 17, 2025 • 0 new comments

• [Core] Ray object primary copy transfer

  #27985 commented on Jun 17, 2025 • 0 new comments

• [Observability] ray timeline errors with ray.rpc.GetAllProfileInfoReply exceeded maximum protobuf size of 2GB

  #27952 commented on Jun 17, 2025 • 0 new comments

• [Core] allow customized error message for WorkerCrashedError

  #27947 commented on Jun 17, 2025 • 0 new comments

• tensorflow.python.framework.errors_impl.NotFoundError: ./multi_worker_model/variables/variables_temp/part-00000-of-00001.index; No such file or directory [Op:MergeV2Checkpoints]

  #27938 commented on Jun 17, 2025 • 0 new comments

• [observability] JSON file generated by Ray timeline doesn't render correctly in the new version of chrome tracing (perfetto)

  #27921 commented on Jun 17, 2025 • 0 new comments

• [RLlib] make policy evaluation support Attention nets

  #27909 commented on Jun 17, 2025 • 0 new comments

• [tune] allow using (nested) data classes for search space definition

  #27904 commented on Jun 17, 2025 • 0 new comments

• [Autoscaler][GCP] Autofill GCP node type resources

  #27888 commented on Jun 17, 2025 • 0 new comments

• [Core] Is it possible to do asynchroneous task submission?

  #29039 commented on Jun 17, 2025 • 0 new comments

• [doc][core] multiprocessing.Pool should document resource usage with ray_remote_args

  #29004 commented on Jun 17, 2025 • 0 new comments

• [Train] Allow passing in placement group to trainer

  #28924 commented on Jun 17, 2025 • 0 new comments

• [<Algorithm overview>]

  #28915 commented on Jun 17, 2025 • 0 new comments

• Runtime Environment Dependencies- container per task

  #28875 commented on Jun 17, 2025 • 0 new comments

• [Ray component: Core] Returning to much data from ray remote fails with no error

  #28855 commented on Jun 17, 2025 • 0 new comments

• Issue on page /ray-core/examples/plot_parameter_server.html

  #28854 commented on Jun 17, 2025 • 0 new comments

• [Datasets] Why does pydantic make training slower？

  #28836 commented on Jun 17, 2025 • 0 new comments

• [Infra] Improve Ray client usability

  #28790 commented on Jun 17, 2025 • 0 new comments

• [Core] Download Logs from Ray Dashboard

  #28788 commented on Jun 17, 2025 • 0 new comments

• Ray Core: AttributeError: 'NoneType' object has no attribute 'enum_types_by_name'

  #28779 commented on Jun 17, 2025 • 0 new comments

• [Tune] HyperOptSearch fails with nested config dicts and points_to_evaluate

  #28753 commented on Jun 17, 2025 • 0 new comments

• Ray Deployment crashes in docker [<Ray component: Serve>]

  #28732 commented on Jun 17, 2025 • 0 new comments

• [Ray Serve]: Testing out on local using Docker container

  #28692 commented on Jun 17, 2025 • 0 new comments

• [core] Generator task that returns more values than specified by num_returns should throw error instead

  #28689 commented on Jun 17, 2025 • 0 new comments

• [Core] CloudPickle explain tool

  #28585 commented on Jun 17, 2025 • 0 new comments

• [dashboard] Dashboard randomly not showing the status of worker nodes.

  #28569 commented on Jun 17, 2025 • 0 new comments

• [AIR] Status updates still prints even with breakpoint

  #28554 commented on Jun 17, 2025 • 0 new comments

• [AIR/Tune] Session report does not show the key for those not included in the first metrics report

  #28549 commented on Jun 17, 2025 • 0 new comments

• [Core] dump the info and anaylze the data offline

  #28496 commented on Jun 17, 2025 • 0 new comments

• [Core] Document what are the generic python code that's easily scalable.

  #28487 commented on Jun 17, 2025 • 0 new comments

• [AIR] Refactor checkpoint encoding and decoding out of Backend to framework-specific Checkpoints

  #28462 commented on Jun 17, 2025 • 0 new comments

• [RLlib][Bug] RLLib Dreamer tuned example requesting unreasonable amount of GPU memory

  #23479 commented on Jun 16, 2025 • 0 new comments

• [Core] Add a warning message if options / arguments differ for Actor.options(get_if_exists=True)

  #23455 commented on Jun 16, 2025 • 0 new comments

• [RLlib][Feature] Feature Importance Plots

  #23447 commented on Jun 16, 2025 • 0 new comments

• [air] If you kill train via control C, a bunch of random error messages show up next time you run Train.

  #23431 commented on Jun 16, 2025 • 0 new comments

• [air] Logging message is not relevant to user

  #23430 commented on Jun 16, 2025 • 0 new comments

• [RLlib][docs] Adding more flow charts to RLlib components docs

  #23393 commented on Jun 16, 2025 • 0 new comments

• [runtime env] Warn user if pip check fails

  #23335 commented on Jun 16, 2025 • 0 new comments

• [runtime env] Refactor packaging code

  #23257 commented on Jun 16, 2025 • 0 new comments

• [Feature] Cleanup current use of `other_args_to_resolve` that passes deployment object into ClassNode

  #23243 commented on Jun 16, 2025 • 0 new comments

• [runtime env] Improve tracking of URI size

  #23186 commented on Jun 16, 2025 • 0 new comments

• [updater][Bug] update fails on preempted node and autoscaler stops scheduling

  #23182 commented on Jun 16, 2025 • 0 new comments

• Pipeline ingress requires trailing /

  #23048 commented on Jun 16, 2025 • 0 new comments

• Shouldn't require `PipelineInputNode` to build a pipeline DAG

  #23037 commented on Jun 16, 2025 • 0 new comments

• Pipeline DAG sanity check for model wrappers fields

  #23019 commented on Jun 16, 2025 • 0 new comments

• Pipeline doesn't accept importable class as arguments

  #23016 commented on Jun 16, 2025 • 0 new comments

• [tune][Bug] 'tune.report( mean_accuracy=sklearn.metrics.accuracy_score(test_y, pred_labels), done=True)'where can i get the mean_accuary result?

  #22992 commented on Jun 16, 2025 • 0 new comments

• [Train] add logging to `finish_training` for existing `Callback`s

  #22754 commented on Jun 16, 2025 • 0 new comments

• [Bug] [serve] Accessing shared objects within a deployment

  #22751 commented on Jun 16, 2025 • 0 new comments

• [Feature] Client version check on commit

  #22675 commented on Jun 16, 2025 • 0 new comments

• [Jobs] run all doc examples in CI

  #22487 commented on Jun 16, 2025 • 0 new comments

• Some tests misusing assertTrue for comparisons

  #22395 commented on Jun 16, 2025 • 0 new comments

• [Enhancement][client] Move synchronous GetObject calls to datapath

  #22357 commented on Jun 16, 2025 • 0 new comments

• [datasets] `random_shuffle` overspills objects on random node

  #17612 commented on Jun 16, 2025 • 0 new comments

• [AIR] `Result` object doesn't work with Ray Client

  #24396 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Current Implementation of Replay Buffer is not a True Circular Buffer

  #24393 commented on Jun 16, 2025 • 0 new comments

• [RLlib] wrong env step counting when train multi-agent with shared default policy

  #24340 commented on Jun 16, 2025 • 0 new comments

• [Core][observability] Enable observability features built in gRPC

  #24327 commented on Jun 16, 2025 • 0 new comments

• [Autoscaler][Docs] Add up-to-date docs on how the autoscaler works.

  #24323 commented on Jun 16, 2025 • 0 new comments

• [Ray Serve Autoscaling] Add release test that checks that nodes scale down when there are no requests

  #24315 commented on Jun 16, 2025 • 0 new comments

• Received message larger than max (105683136 vs. 104857600)

  #24286 commented on Jun 16, 2025 • 0 new comments

• [Serve] [Doc] HTTP Adapters Cookbooks

  #24245 commented on Jun 16, 2025 • 0 new comments

• [Serve] Default DAGDriver implementation cannot serve.run() or serve.build() twice

  #24122 commented on Jun 16, 2025 • 0 new comments

• [AIR] Support functionality to stitch Preprocessor with Keras model

  #24023 commented on Jun 16, 2025 • 0 new comments

• [Core] Log propagation between actor exit called and process terminated

  #24020 commented on Jun 16, 2025 • 0 new comments

• [<Ray component: Serve] Improve access by index/key on intermediate result in Serve deployment graph

  #23987 commented on Jun 16, 2025 • 0 new comments

• [Serve] [Docs] Improve architectural diagrams

  #23956 commented on Jun 16, 2025 • 0 new comments

• [Runtime Env] Dependency Installation private git repositories via ssh

  #23768 commented on Jun 16, 2025 • 0 new comments

• [ray client] ray.wait timeout is not respected when connection is interrupted

  #23694 commented on Jun 16, 2025 • 0 new comments

• [Feature] [Tune] Trial-wise dependencies

  #23654 commented on Jun 16, 2025 • 0 new comments

• [Bug] `policies_to_train` throws incorrect/confusing error message when passed an empty list.

  #23646 commented on Jun 16, 2025 • 0 new comments

• [Feature] support of complicated action space in QMix algorithm in Rllib.

  #23634 commented on Jun 16, 2025 • 0 new comments

• [runtime env] Deflake `test_runtime_env_working_dir_2`

  #23569 commented on Jun 16, 2025 • 0 new comments

• [runtime env] [Feature] Make Internal KV operations async

  #23567 commented on Jun 16, 2025 • 0 new comments

• [Feature] .bind() on function does not take pre-bind value from upstream DAGNode

  #23511 commented on Jun 16, 2025 • 0 new comments

• [Tune] [Bug] Ray checkpoint sync can sometimes fail to upload checkpoints to s3, plus log spew about sync client observed

  #21469 commented on Jun 16, 2025 • 0 new comments

• [runtime env] raise exception for unsupported runtime_env features on Windows

  #21435 commented on Jun 16, 2025 • 0 new comments

• [train] fix scalability of `JsonLoggerCallback`

  #21416 commented on Jun 16, 2025 • 0 new comments

• [Feature] [runtime env] [java] select jdk version

  #21239 commented on Jun 16, 2025 • 0 new comments

• [Feature][Tune] Trial status based Stopper

  #21222 commented on Jun 16, 2025 • 0 new comments

• [Train][Tune] Unify Train and Tune Callbacks

  #21065 commented on Jun 16, 2025 • 0 new comments

• [Bug] rsync_filter isn't used in hash_runtime_conf

  #20878 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] Persistent problems encountered during autoscaling can lead to driver log spam

  #20855 commented on Jun 16, 2025 • 0 new comments

• [Bug] Test placement group chaos testing

  #20716 commented on Jun 16, 2025 • 0 new comments

• [GCP][autoscaler] Scale down is slow and Ray status doesn't show pending nodes

  #20695 commented on Jun 16, 2025 • 0 new comments

• Support snappy compression for spilled objects

  #20575 commented on Jun 16, 2025 • 0 new comments

• Sparse object reads - read part of an object, without downloading the entire object

  #20500 commented on Jun 16, 2025 • 0 new comments

• [core] Scale shuffle to 200+ nodes

  #20499 commented on Jun 16, 2025 • 0 new comments

• Memory-aware task scheduling to avoid OOMs under memory pressure

  #20495 commented on Jun 16, 2025 • 0 new comments

• [Feature] [Placement Group] Add timeout mechanism when scheduling placement group

  #20477 commented on Jun 16, 2025 • 0 new comments

• [job submission] Add RAY_ADDRESS or --address to suggested commands for logs/status

  #20441 commented on Jun 16, 2025 • 0 new comments

• [Bug] [Ray Autoscaler] [Core] Ray Worker Node Relaunching during 'ray up'

  #20402 commented on Jun 16, 2025 • 0 new comments

• [workflow] Fail to construct workflow within a workflow

  #20381 commented on Jun 16, 2025 • 0 new comments

• [Feature] [Serve] Threading for Ray Serve

  #20169 commented on Jun 16, 2025 • 0 new comments

• [Feature] [Serve] Support Sticky Sessions for Stateful Workflows Deployed via Ray Serve

  #20107 commented on Jun 16, 2025 • 0 new comments

• [runtime env] Remove filelock dependency

  #20083 commented on Jun 16, 2025 • 0 new comments

• [Client] Dataset write_csv AttributeError: ‘Worker’ object has no attribute 'core_worker'

  #35537 commented on Jun 16, 2025 • 0 new comments

• Enhance state notification pattern in Ray pubsub

  #22340 commented on Jun 16, 2025 • 0 new comments

• [Core] Avoiding subscribing to all logs by each log subscriber

  #22274 commented on Jun 16, 2025 • 0 new comments

• [Train] TPU support

  #22251 commented on Jun 16, 2025 • 0 new comments

• [train] support per epoch shuffling with `prepare_dataloader`

  #22108 commented on Jun 16, 2025 • 0 new comments

• [runtime env] Refactor `pip` protobuf to store a single str (`requirements.txt` contents) instead of list of "packages"

  #22097 commented on Jun 16, 2025 • 0 new comments

• [runtime_env] Remove `.lock` files after URI garbage collection

  #22062 commented on Jun 16, 2025 • 0 new comments

• [runtime env] Use LRU cache for URIs instead of random eviction

  #22060 commented on Jun 16, 2025 • 0 new comments

• [runtime env] Use single URI for `py_modules` field

  #22059 commented on Jun 16, 2025 • 0 new comments

• [Train] Add callback preprocessor that smoothly tracks values

  #21989 commented on Jun 16, 2025 • 0 new comments

• [Bug] Policy - ActionDistribution Type

  #21973 commented on Jun 16, 2025 • 0 new comments

• [runtiime env] Use coroutine to create runtime envs in `runtime_env_agent`

  #21950 commented on Jun 16, 2025 • 0 new comments

• [Train] Add support for Bagua

  #21934 commented on Jun 16, 2025 • 0 new comments

• [Bug] "The kernel has died..." during Ray tune.run

  #21917 commented on Jun 16, 2025 • 0 new comments

• [Jobs] Backwards compatibility tests for REST API

  #21915 commented on Jun 16, 2025 • 0 new comments

• [Jobs] Make jobs work out-of-the-box with cluster YAML

  #21911 commented on Jun 16, 2025 • 0 new comments

• [Train] Support for averaging results

  #21849 commented on Jun 16, 2025 • 0 new comments

• AttributeError raised when using response_model in FastAPI route decorator

  #21744 commented on Jun 16, 2025 • 0 new comments

• [Feature] [runtime env] [C++] support a strong-typed API in C++

  #21733 commented on Jun 16, 2025 • 0 new comments

• [runtime env] Cross-language runtime env

  #21731 commented on Jun 16, 2025 • 0 new comments

• [Testing] multi fake node set up doesn't work under non ray client mode

  #21653 commented on Jun 16, 2025 • 0 new comments

• [Bug] "Sent message larger than max" error with dask

  #21601 commented on Jun 16, 2025 • 0 new comments

• [runtime env] Can we avoid merging two runtime envs?

  #21494 commented on Jun 16, 2025 • 0 new comments

• [Core] Ray Actor abnormal exit problem && Reproduction

  #17198 commented on Jun 16, 2025 • 0 new comments

• Support resizing placement groups

  #16403 commented on Jun 16, 2025 • 0 new comments

• [dashboard] Errors are not shown

  #15238 commented on Jun 16, 2025 • 0 new comments

• changing the docker image in consecutive `ray up` calls fails.

  #14990 commented on Jun 16, 2025 • 0 new comments

• [metrics] Add regression tests for Prometheus metrics

  #14614 commented on Jun 16, 2025 • 0 new comments

• [dashboard] Show more nodes at a time instead of paging through

  #14537 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] Support Memory Aware Scheduling on a multi-node-type cluster.

  #14104 commented on Jun 16, 2025 • 0 new comments

• [RFC] k8s-native worker pool

  #14077 commented on Jun 16, 2025 • 0 new comments

• [dashboard] clicking on a column to sort makes the UI blank

  #13525 commented on Jun 16, 2025 • 0 new comments

• Autoscaler does not respect --num-cpus argument to `ray start`

  #13270 commented on Jun 16, 2025 • 0 new comments

• [core] Number of CPUs in ray.available_resources() does not match Dashboard's Machine View

  #13100 commented on Jun 16, 2025 • 0 new comments

• atexit handlers don't run when actor is terminated from going out of scope

  #12806 commented on Jun 16, 2025 • 0 new comments

• Task Cancellation is broken for queued tasks

  #12080 commented on Jun 16, 2025 • 0 new comments

• [logging] Use 'warnings.warn' appropriately

  #12060 commented on Jun 16, 2025 • 0 new comments

• [Dashboard] New dashboard port errors in a large cluster.

  #11638 commented on Jun 16, 2025 • 0 new comments

• ES Trainer does not support evaluation workers

  #10999 commented on Jun 16, 2025 • 0 new comments

• [Plasma] Improve plasma documentation on distributed storage

  #10858 commented on Jun 16, 2025 • 0 new comments

• Unable to connect to ray head running on linux from ray worker node on windows

  #10362 commented on Jun 16, 2025 • 0 new comments

• Ray log tracing

  #9786 commented on Jun 16, 2025 • 0 new comments

• [Core] Logging policy should be clearly defined and needs unit test coverage

  #9692 commented on Jun 16, 2025 • 0 new comments

• [dashboard] Error on Infinity values

  #9103 commented on Jun 16, 2025 • 0 new comments

• [Feature][Client] remove ray.disconnect() and ray.connect()

  #22125 commented on Jun 16, 2025 • 0 new comments

• [Bug] Detached actor exceptions are not logged.

  #21810 commented on Jun 16, 2025 • 0 new comments

• [Bug] Sometimes the worker node logs in the ray dashboard are empty

  #21785 commented on Jun 16, 2025 • 0 new comments

• [Core][Feature] Add checksum support for object store.

  #21782 commented on Jun 16, 2025 • 0 new comments

• Setting VF_SHARE_LAYERS to False and NO_FINAL_LINEAR to true leads to a bug

  #21756 commented on Jun 16, 2025 • 0 new comments

• [Feature] [runtime env] support using different python versions in Ray cluster

  #21597 commented on Jun 16, 2025 • 0 new comments

• [Feature] [Serve] Request Redistribution Among Replicas

  #21578 commented on Jun 16, 2025 • 0 new comments

• Put failed error occurred when shutdown and init again at client mode

  #21573 commented on Jun 16, 2025 • 0 new comments

• [Core] [Bug] No timeout or deadlock on scheduling job in remote cluster

  #21419 commented on Jun 16, 2025 • 0 new comments

• [Feature] [Autoscaler] Scaling Intelligently Based on Observed Resource Bottlenecks (related: task & actor profiling)

  #21301 commented on Jun 16, 2025 • 0 new comments

• [Core] [Bug] Failed to register worker to Raylet for single node, multi-GPU

  #21226 commented on Jun 16, 2025 • 0 new comments

• [Train] Port over `timm` example to Train

  #21020 commented on Jun 16, 2025 • 0 new comments

• [Bug] [RLlib] Custom metrics are not reported to Tune

  #20938 commented on Jun 16, 2025 • 0 new comments

• [Train] Deepspeed support

  #20648 commented on Jun 16, 2025 • 0 new comments

• [Bug] Cannot start cluster if other user is already running one

  #20634 commented on Jun 16, 2025 • 0 new comments

• [Bug] Excess memory usage when scheduling tasks in parallel?

  #20618 commented on Jun 16, 2025 • 0 new comments

• [Bug] Ray auto init interacts badly with allow_multiple=True and kills python shell

  #20355 commented on Jun 16, 2025 • 0 new comments

• [Bug] BasicVariantGenerator not compatible with Repeater

  #19879 commented on Jun 16, 2025 • 0 new comments

• [Feature] Support Sigopt for Tune standard space definitions

  #19018 commented on Jun 16, 2025 • 0 new comments

• [Bug] Re-enable Worker in Container Tests.

  #18787 commented on Jun 16, 2025 • 0 new comments

• [GCP][autoscaler] Rework ray TPU demos to create nothing but TPU VMs (no harddrives / `n2-standard-2` instances)

  #18645 commented on Jun 16, 2025 • 0 new comments

• Ray kill actor API is a GET request

  #18411 commented on Jun 16, 2025 • 0 new comments

• [AIR/Train] Torch: Automatically unpack model when checkpointing state dicts

  #24975 commented on Jun 16, 2025 • 0 new comments

• [AIR/Train] Automatically return the framework specific dataset in `train_loop_per_worker`

  #24974 commented on Jun 16, 2025 • 0 new comments

• [<Ray component: RLlib>] ppo error when not using critic

  #24907 commented on Jun 16, 2025 • 0 new comments

• [RLlib]: Add tabular models to ModelV2

  #24882 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Error when converting GYM Robotics env to Multi-agent Env with the make_multi_agent wrapper

  #24881 commented on Jun 16, 2025 • 0 new comments

• [tune] SigOptSearch suggester is not serialisable

  #24864 commented on Jun 16, 2025 • 0 new comments

• [core] Add basic metrics for lineage reconstruction

  #24855 commented on Jun 16, 2025 • 0 new comments

• [Core] Enhance runtime env state when `ray list runtime-env` is used.

  #24838 commented on Jun 16, 2025 • 0 new comments

• [Core] Refactor Ray memory codepath to follow same pattern as `ray list tasks`.

  #24836 commented on Jun 16, 2025 • 0 new comments

• [Core] Reach parity of task status for `ray memory` and `ray list tasks`

  #24835 commented on Jun 16, 2025 • 0 new comments

• [Tune] `MedianStoppingRule` mishandles `nan`s

  #24809 commented on Jun 16, 2025 • 0 new comments

• [Serve] Simplify json_serde of deployment graph

  #24620 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Metrics not reported with Client/Server and env=None

  #24601 commented on Jun 16, 2025 • 0 new comments

• [Core] In C++, there are D_GLIBCXX_USE_CXX11_ABI settings conflicts when both Ray and Arrow are used.

  #24566 commented on Jun 16, 2025 • 0 new comments

• [Serve] `Deployment.url` not updated after options changing name or prefix.

  #24548 commented on Jun 16, 2025 • 0 new comments

• [Rllib] Lack validation for "num_workers" parameter in DDPGTrainer.

  #24536 commented on Jun 16, 2025 • 0 new comments

• [doc] Update instructions for wheel installation

  #24533 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Simplex action space shape

  #24529 commented on Jun 16, 2025 • 0 new comments

• [Tune] Make it easy to configure logger level

  #24447 commented on Jun 16, 2025 • 0 new comments

• [tune] improve documentation around "resource exhausted error"

  #24439 commented on Jun 16, 2025 • 0 new comments

• [Core] Unify RegisterClient and AnnounceWorkerPort

  #24432 commented on Jun 16, 2025 • 0 new comments

• [core] Annotation and docstring for ray.remote wrapped functions

  #24411 commented on Jun 16, 2025 • 0 new comments

• "ray timeline" command fails when RAY_ADDRESS is set

  #8951 commented on Jun 16, 2025 • 0 new comments

• [tune] [dashboard] Table formatting issues due to too many hparams

  #8667 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] "Cannot perform an interactive login from a non TTY device" when trying to use a private docker registry

  #7339 commented on Jun 16, 2025 • 0 new comments

• Allowing multiple users to access a single ray cluster

  #6800 commented on Jun 16, 2025 • 0 new comments

• [Ray core & ray cluster] Add diagrams/architectures to explain how to run ray locally vs remotely

  #25663 commented on Jun 16, 2025 • 0 new comments

• [Ray Clusters] Remove nightly and latest images and wheels from all example configs.

  #25606 commented on Jun 16, 2025 • 0 new comments

• [air] Consider having a preprocessor for Feast integration

  #25559 commented on Jun 16, 2025 • 0 new comments

• [Core] Open telemetry Context pass from ray client to actors

  #25538 commented on Jun 16, 2025 • 0 new comments

• [dataset] Reduce tasks in push-based shuffle are not evenly distributed

  #25468 commented on Jun 16, 2025 • 0 new comments

• [Core] [State Observability] List all actor logs when actors are restarted.

  #25443 commented on Jun 16, 2025 • 0 new comments

• [air] Ordinal Encoder complains about None

  #25442 commented on Jun 16, 2025 • 0 new comments

• [Autoscaler] google-cloud-storage seems cannot read GOOGLE_APPLICATION_CREDENTIALS

  #25308 commented on Jun 16, 2025 • 0 new comments

• [Serve] Dynamically move models between CPUs and GPUs

  #25295 commented on Jun 16, 2025 • 0 new comments

• [RLlib][Doc] Add documentation for `ModelCatalog.get_model_v2()`

  #25186 commented on Jun 16, 2025 • 0 new comments

• [AIR] MLflow integration polish

  #25156 commented on Jun 16, 2025 • 0 new comments

• [AIR] TensorFlow warns to use `distribute.MultiWorkerMirroredStrategy` when I'm already using it

  #25140 commented on Jun 16, 2025 • 0 new comments

• [air] Have a default column for not frequent enough categories for OHE

  #25096 commented on Jun 16, 2025 • 0 new comments

• [Core] Make NodeManager unit testable

  #25095 commented on Jun 16, 2025 • 0 new comments

• [AIR] Improve logging for train

  #25088 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Hope RLlib can support DQfD & POfD

  #25058 commented on Jun 16, 2025 • 0 new comments

• [AIR] Support postprocessing in Predictors

  #24979 commented on Jun 16, 2025 • 0 new comments

• [AIR] Add a `TorchVision` preprocessor

  #24976 commented on Jun 16, 2025 • 0 new comments

• [Fix][GCS] Implement reconnection for RedisContext

  #48781 commented on Jun 22, 2025 • 0 new comments

• [Core]: Fix ConnectionError on Autoscaler CR lookups in K8s clusters …

  #48675 commented on Jun 22, 2025 • 0 new comments

• [runtime env]: Integrating ROCm Systems Profiler to Ray worker process

  #48525 commented on Jun 19, 2025 • 0 new comments

• Fix invalid type for progress_reporter parameter of RunConfig

  #48439 commented on Jun 19, 2025 • 0 new comments

• [doc] fix: Typo and missing import in doc

  #48311 commented on Jun 19, 2025 • 0 new comments

• [WIP][core] C++20 upgrade

  #48044 commented on Jun 19, 2025 • 0 new comments

• :bug: do not modify user-provided runtime_env

  #48021 commented on Jun 22, 2025 • 0 new comments

• [Data] Fix parallelism deriving heuristic to ensure parallelism stays w/in min/max bounds

  #47695 commented on Jun 20, 2025 • 0 new comments

• Add generic item support for queue

  #46849 commented on Jun 18, 2025 • 0 new comments

• blind try on ubuntu upgrade ..

  #45427 commented on Jun 16, 2025 • 0 new comments

• [core] object store data transfer zstd

  #44755 commented on Jun 17, 2025 • 0 new comments

• [data] add better support for list-typed fields when using `write_bigquery`

  #44564 commented on Jun 22, 2025 • 0 new comments

• Ray IPv6 support

  #44252 commented on Jun 19, 2025 • 0 new comments

• Adapt the joblib backend for compatibility with `return_as=generator`

  #41028 commented on Jun 16, 2025 • 0 new comments

• [dashboard] ignore reinit error when getting dashboard url

  #40545 commented on Jun 19, 2025 • 0 new comments

• Update pettingzoo_env.py

  #39431 commented on Jun 17, 2025 • 0 new comments

• [ci] remove is_automated_build in setup.py

  #36547 commented on Jun 16, 2025 • 0 new comments

• [RLLib][Air] MLFlow parsing of RLLib evaluation and custom metrics

  #26711 commented on Jun 17, 2025 • 0 new comments

• [Serve] Make replica scheduler backoff configurable

  #52871 commented on Jun 21, 2025 • 0 new comments

• [core|serve] Migrate shared utilities from `ray._private` to `ray._common`

  #53478 commented on Jun 21, 2025 • 0 new comments

• CI test linux://rllib:examples/metrics/custom_metrics_in_algorithm_training_step is flaky

  #51870 commented on Jun 21, 2025 • 0 new comments

• [core] Cleanup gcs event listeners and gcs_storage env variable

  #53566 commented on Jun 19, 2025 • 0 new comments

• update to protbuf-28.2, absl-20240722, grpc-1.67 and patch for windows

  #51673 commented on Jun 20, 2025 • 0 new comments

• [Docs][wip] Feature: adopt llms.txt convention

  #51605 commented on Jun 20, 2025 • 0 new comments

• [Refactor]Rename NCCL-related items to comm_backend

  #51061 commented on Jun 20, 2025 • 0 new comments

• [doc] add jax example

  #51040 commented on Jun 22, 2025 • 0 new comments

• Suppress type error

  #50994 commented on Jun 22, 2025 • 0 new comments

• fix restore BUG "RuntimeError: Expected scalars to be on CPU, got cud…

  #50983 commented on Jun 22, 2025 • 0 new comments

• [CI] Enable pretty-format-java pre-commit hook

  #50957 commented on Jun 22, 2025 • 0 new comments

• [core] Cover cpplint for ray/src/ray/stats

  #50678 commented on Jun 22, 2025 • 0 new comments

• [Core] Split stats_metric into smaller targets to improve build performance

  #50595 commented on Jun 22, 2025 • 0 new comments

• [RLlib] Enable spliting and zero padding of Dict observation

  #50589 commented on Jun 22, 2025 • 0 new comments

• [Autoscaler][V2] Use running node instances to rate-limit upscaling

  #50414 commented on Jun 22, 2025 • 0 new comments

• [core][collective] Avoid creation of `gloo_queue` in race condition

  #50132 commented on Jun 22, 2025 • 0 new comments

• Update multi-agent-envs.rst

  #50075 commented on Jun 22, 2025 • 0 new comments

• [core] Thread-safe gcs node manager

  #50024 commented on Jun 22, 2025 • 0 new comments

• [core] Don't get dashboard address after each dashboard connection failure

  #49584 commented on Jun 21, 2025 • 0 new comments

• [core][cgraph] Use cv instead of busy wait for next version

  #49542 commented on Jun 21, 2025 • 0 new comments

• [RLlib] Add NPU and HPU support to RLlib

  #49535 commented on Jun 17, 2025 • 0 new comments

• [core][cgraph] Use threadpool and one io_context for mutable object provider

  #49500 commented on Jun 21, 2025 • 0 new comments

• [Fix][Core] Periodically check log message queue cleared before shutdown

  #49337 commented on Jun 22, 2025 • 0 new comments

• [WIP] Remove VM cluster autoscaler docker implementation

  #49238 commented on Jun 17, 2025 • 0 new comments

• [train] Make dataset argument covariant

  #48999 commented on Jun 19, 2025 • 0 new comments

• [flaky] test_scheduling_2.py::test_demand_report_when_scale_up

  #53811 commented on Jun 19, 2025 • 0 new comments

• [Serve] Allow HTTPs Options in Ray Serve

  #26814 commented on Jun 19, 2025 • 0 new comments

• [Core] Transient network failure on RPC `WaitForActorRefDeleted` causes actor registration fail

  #53797 commented on Jun 18, 2025 • 0 new comments

• [Core] Transient network failure on RPC `MarkJobFinished` causes node crash

  #53645 commented on Jun 18, 2025 • 0 new comments

• Core: Ray 2.45 causes Google's LIBTPU to be very spammy

  #53756 commented on Jun 18, 2025 • 0 new comments

• [Core] Make Ray Core tasks/actors metrics counters (accumulators)

  #47522 commented on Jun 18, 2025 • 0 new comments

• [RLlib]

  #52683 commented on Jun 18, 2025 • 0 new comments

• CI test linux://python/ray/tests:test_runtime_env_container is consistently_failing

  #45223 commented on Jun 18, 2025 • 0 new comments

• [Core] Ray causes a 25% slower GPU performance compared with manually written Multi-processing program on 8 Hopper GPUs

  #53799 commented on Jun 18, 2025 • 0 new comments

• [Core] ASSERTION FAILED: queue.num_items() == 0

  #53510 commented on Jun 18, 2025 • 0 new comments

• [Serve] Proxy issues: Request cancellation, intermittent 503 backpressure, and max_queued_requests configuration not applied

  #53794 commented on Jun 18, 2025 • 0 new comments

• [Dashboard] Discrepancy between Worker Process Memory Display on Dashboard and RSS Statistics

  #53829 commented on Jun 18, 2025 • 0 new comments

• [<Ray component: Core|RLlib|etc...>] Issue of port allocation

  #53790 commented on Jun 18, 2025 • 0 new comments

• [Dashboard] Support ncu

  #53759 commented on Jun 18, 2025 • 0 new comments

• Incorrect default value of CUBLAS_WORKSPACE_CONFIG

  #47690 commented on Jun 18, 2025 • 0 new comments

• [Serve] make various default values of `AutoscalingConfig.max_replicas` consistent and >1

  #50222 commented on Jun 18, 2025 • 0 new comments

• CI test linux://rllib:learning_tests_stateless_cartpole_appo_gpu is flaky

  #47295 commented on Jun 18, 2025 • 0 new comments

• [core][compiled graph] Support all-to-one collective ops (e.g. reduce)

  #49324 commented on Jun 18, 2025 • 0 new comments

• [autoscaler] SubnetId, a valid AWS field, is being ignored in cluster yaml

  #14551 commented on Jun 18, 2025 • 0 new comments

• [Core] The streaming generator will mark the inplasma object that is already ready as failed after the task fails.

  #53772 commented on Jun 17, 2025 • 0 new comments

• [core] control whether to construct a default concurrency group executor when max-concurrency=1 and there are other concurrency group for an actor

  #53771 commented on Jun 17, 2025 • 0 new comments

• [core] Race condition between raylet graceful shutdown and GCS health checks

  #53739 commented on Jun 17, 2025 • 0 new comments

• [Core] BUG: Cluster crashes when using temp_dir "could not connect to socket" raylet.x [since 2.7+]

  #44431 commented on Jun 20, 2025 • 0 new comments

• [Ray debugger] Unable to use debugger on slurm cluster

  #51157 commented on Jun 20, 2025 • 0 new comments

• [Core] Submitted containerized job is stuck in pending mode

  #37293 commented on Jun 20, 2025 • 0 new comments

• [Data] Custom Partitioner in Ray Data and Related Implementation Considerations

  #53800 commented on Jun 20, 2025 • 0 new comments

• [core][gpu-objects] Ability to register custom types for GPU data

  #52340 commented on Jun 20, 2025 • 0 new comments

• CI test linux://rllib:learning_tests_multi_agent_pendulum_sac_multi_cpu is consistently_failing

  #47264 commented on Jun 19, 2025 • 0 new comments

• [data] Bad error message when function outputs cannot be pickled

  #46642 commented on Jun 19, 2025 • 0 new comments

• [data] ObjectRefs passed to map UDF are not automatically deref'ed

  #49207 commented on Jun 19, 2025 • 0 new comments

• [Core] Ray hangs with vllm0.8.5 v1 api for tp8+pp4

  #53758 commented on Jun 19, 2025 • 0 new comments

• [data] Optimize Dataset.unique()

  #38764 commented on Jun 19, 2025 • 0 new comments

• Error Handling Large Pyarrow Chunk

  #53536 commented on Jun 19, 2025 • 0 new comments

• [<Ray component: Data>] lack of check for empty table produce lots of error messages

  #53605 commented on Jun 19, 2025 • 0 new comments

• Release test random_shuffle_fixed_size failed

  #53806 commented on Jun 19, 2025 • 0 new comments

• [Data] Support for SQL/DataFrame capability

  #53693 commented on Jun 19, 2025 • 0 new comments

• [RayData] The write operator supports the use of an actor pool

  #53552 commented on Jun 19, 2025 • 0 new comments

• [core][gpu-objects] Actor sends the same ObjectRef twice to another actor

  #51273 commented on Jun 19, 2025 • 0 new comments

• [Autoscaler] Improve NodeProvider interface, make it easier to extend it to cluster managers (e.g. Fargate)

  #25134 commented on Jun 19, 2025 • 0 new comments

• [RLlib] Observation space with 2 dimensions not working with the new API stack

  #46631 commented on Jun 19, 2025 • 0 new comments

• [Ray Client] - Client server failed with runtime_env container

  #29852 commented on Jun 19, 2025 • 0 new comments

• [Ray Train] XGBoostTrainer crashes with ActorDiedError when using num_workers > 1 and use_gpu=False

  #53123 commented on Jun 19, 2025 • 0 new comments

• CI test windows://python/ray/serve/tests:test_request_timeout is flaky

  #48417 commented on Jun 19, 2025 • 0 new comments

• CI test linux://rllib:examples/algorithms/appo_custom_algorithm_w_shared_data_actor is flaky

  #53176 commented on Jun 19, 2025 • 0 new comments

• Bump torch from 2.3.0 to 2.7.1 in /python

  #53558 commented on Jun 19, 2025 • 0 new comments

• Script to generate test coverage for doc files

  #53556 commented on Jun 16, 2025 • 0 new comments

• [core] Support pip_install_options for pip

  #53551 commented on Jun 20, 2025 • 0 new comments

• [RLlib] Upgrade RLlink protocol for external env/simulator training.

  #53550 commented on Jun 17, 2025 • 0 new comments

• [data] add Lance-based ordered data conversion that keeps row_id content unchanged

  #53542 commented on Jun 19, 2025 • 0 new comments

• [serve.llm] Update ray-llm docker

  #53532 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Wrapper which allows EnvRunners to operate on environments with Repeated observation spaces

  #53519 commented on Jun 19, 2025 • 0 new comments

• [core][telemetry/09] record sum metric e2e

  #53512 commented on Jun 20, 2025 • 0 new comments

• [Dashboard] Fixing residual state leaks in Dashboard/Agent

  #53508 commented on Jun 20, 2025 • 0 new comments

• [Data] Add a data compaction function

  #53489 commented on Jun 19, 2025 • 0 new comments

• [WIP][Data] Batch query for block_ref_iter

  #53485 commented on Jun 18, 2025 • 0 new comments

• docs test coverage script

  #53482 commented on Jun 18, 2025 • 0 new comments

• [core][compiled graphs] Unify and simplify NCCL operation nodes

  #53470 commented on Jun 21, 2025 • 0 new comments

• [Data] Add dropna function

  #53464 commented on Jun 18, 2025 • 0 new comments

• [Serve] Set the docs path after app is initialized on the replica

  #53463 commented on Jun 22, 2025 • 0 new comments

• [core] Check if a task can be spilled before checking if args can be pinned

  #53462 commented on Jun 19, 2025 • 0 new comments

• [Data] Add fillna function

  #53459 commented on Jun 18, 2025 • 0 new comments

• [WIP][Data] Add support for Arrow native fixed-shape tensor type

  #53450 commented on Jun 18, 2025 • 0 new comments

• Bump torch from 2.0.1 to 2.7.0 in /doc/source/templates/testing/docker/03_serving_stable_diffusion

  #53447 commented on Jun 19, 2025 • 0 new comments

• [Data] add switch for optimizer rules

  #53427 commented on Jun 20, 2025 • 0 new comments

• [Data] Add support for ray.dataset.map_sql

  #53417 commented on Jun 20, 2025 • 0 new comments

• [Doc][KubeRay] remove head pod trailing hash and adjust volcano output

  #53826 commented on Jun 16, 2025 • 0 new comments

• [core] Add switch for the cache of runtime env

  #53775 commented on Jun 21, 2025 • 0 new comments

• [Core][Bug fix]Fix issue: the streaming generator will mark the inplasma object that is already ready as failed after the task fails.

  #53773 commented on Jun 17, 2025 • 0 new comments

• [core] Control whether to construct a default concurrency group executor when max-concurrency=1 and there are other concurrency groups for an actor

  #53770 commented on Jun 17, 2025 • 0 new comments

• [core] upgrade opentelemetry-sdk

  #53745 commented on Jun 20, 2025 • 0 new comments

• [core][telemetry/11] support histogram metric on worker side

  #53740 commented on Jun 20, 2025 • 0 new comments

• [core][telemetry/10] support custom gauge+counter+sum metrics

  #53734 commented on Jun 20, 2025 • 0 new comments

• (serve.llm) Remove test leakage from placement bundle logic

  #53723 commented on Jun 20, 2025 • 0 new comments

• [RLlib; Offline RL] Implement Offline Policy Evaluation (OPE) via Importance Sampling.

  #53702 commented on Jun 20, 2025 • 0 new comments

• [Data] Add reading from Delta Lake tables and from Unity Catalog

  #53701 commented on Jun 21, 2025 • 0 new comments

• [data] allow max_calls to be a static but not dynamic option

  #53687 commented on Jun 19, 2025 • 0 new comments

• [Serve] Check multiple FastAPI ingress deployments in a single application

  #53647 commented on Jun 22, 2025 • 0 new comments

• [core] Gcs actor manager cleanup

  #53633 commented on Jun 22, 2025 • 0 new comments

• [rllib] IMPALA fix no attribute '_minibatch_size'

  #53620 commented on Jun 21, 2025 • 0 new comments

• [serve.llm] Add useful logging in prefill_decode_disagg.py

  #53604 commented on Jun 16, 2025 • 0 new comments

• [core] Cleanup retryable grpc client

  #53599 commented on Jun 21, 2025 • 0 new comments

• [Do not merge] Run ray data release tests with export API

  #53594 commented on Jun 21, 2025 • 0 new comments

• BLD: Update ``.bazelrc`` file for Windows 11 build

  #53586 commented on Jun 20, 2025 • 0 new comments

• [CI] Re-enable isort for all remaining files

  #53583 commented on Jun 22, 2025 • 0 new comments

• [Not for Merge] Event Aggregator Perf

  #53576 commented on Jun 19, 2025 • 0 new comments

• [DON'T MERGE]

  #53575 commented on Jun 20, 2025 • 0 new comments

• [Data] [Draft] user guide for aggregations

  #53568 commented on Jun 21, 2025 • 0 new comments

• [data] New landing page with better examples that show key workloads

  #53228 commented on Jun 17, 2025 • 0 new comments

• [core] enable -Wshadow for all c++ targets

  #53194 commented on Jun 17, 2025 • 0 new comments

• [core] Returning a useful message when trying to get logs for a job that has not started yet

  #53174 commented on Jun 17, 2025 • 0 new comments

• [draft] Submit Ray release test as RayJob to Kuberay GKE

  #53165 commented on Jun 20, 2025 • 0 new comments

• [data] fix lance count_rows not support filter

  #53162 commented on Jun 17, 2025 • 0 new comments

• [core] Don't try to monitor zipped files

  #53151 commented on Jun 19, 2025 • 0 new comments

• Make vllm_engine a deployment

  #53139 commented on Jun 17, 2025 • 0 new comments

• Fix broken Ray Workflows documentation link in README.rst

  #53136 commented on Jun 17, 2025 • 0 new comments

• feat(runtime_env): add Azure Blob Storage support

  #53135 commented on Jun 17, 2025 • 0 new comments

• macos wheel build debug

  #53119 commented on Jun 17, 2025 • 0 new comments

• Bump flask-cors from 4.0.0 to 6.0.0 in /python

  #53116 commented on Jun 17, 2025 • 0 new comments

• WIP: Add iter_torch_batches Tensor cache

  #53069 commented on Jun 19, 2025 • 0 new comments

• [core] Remove tests that are permanently skipped with old decorator

  #53046 commented on Jun 17, 2025 • 0 new comments

• [core] Node manager related cpp cleanup

  #52990 commented on Jun 17, 2025 • 0 new comments

• [RLlib; Offline RL] - Use `iter_torch_batches` in learner

  #52968 commented on Jun 20, 2025 • 0 new comments

• [Data] fix write_iceberg error

  #52956 commented on Jun 17, 2025 • 0 new comments

• [Core] Ensure Ray vendored libraries only be visible and used by Ray internal

  #52905 commented on Jun 17, 2025 • 0 new comments

• [Data] added XML datasource

  #52539 commented on Jun 17, 2025 • 0 new comments

• [core] Minor task manager related improvements

  #52294 commented on Jun 19, 2025 • 0 new comments

• [Core] Deserialization of PyArrow Extension Arrays by registration of deserializers

  #51972 commented on Jun 20, 2025 • 0 new comments

• Add new autoscaling parameter `aggregation function`

  #51905 commented on Jun 19, 2025 • 0 new comments

• [core] Remove client call tag

  #51817 commented on Jun 19, 2025 • 0 new comments

• kuberay edits

  #53411 commented on Jun 18, 2025 • 0 new comments

• [WIP] Remove `_owner` arg for `ray.put`

  #53410 commented on Jun 18, 2025 • 0 new comments

• [Core] Add Logic to Emit Task Events to Event Aggregator

  #53402 commented on Jun 19, 2025 • 0 new comments

• [serve.llm] DO NOT REVIEW, IN DRAFT

  #53391 commented on Jun 16, 2025 • 0 new comments

• Deduplicate schema in BlockMetadata

  #53384 commented on Jun 18, 2025 • 0 new comments

• [serve.llm] [cleanup] Add LLMConfig.parse_from() api

  #53382 commented on Jun 18, 2025 • 0 new comments

• Bump vllm from 0.8.5 to 0.9.0 in /python

  #53375 commented on Jun 18, 2025 • 0 new comments

• [WIP] Remove global worker

  #53372 commented on Jun 18, 2025 • 0 new comments

• Filter out ANSI escape codes from logs when retrieving logs from the dashboard

  #53370 commented on Jun 19, 2025 • 0 new comments

• fix: Type of AlgorithmConfig.training(learner_connector

  #53369 commented on Jun 19, 2025 • 0 new comments

• [core] Cleanup plasma client and object manager

  #53357 commented on Jun 19, 2025 • 0 new comments

• [Docs] Clarify Train-side docs on Ray Data

  #53349 commented on Jun 20, 2025 • 0 new comments

• [RLlib] ConnectorV2 API polishings (stricter input-/output batch formats).

  #53328 commented on Jun 18, 2025 • 0 new comments

• [core] Avoid making rpc for local GetLocationFromOwner

  #53322 commented on Jun 17, 2025 • 0 new comments

• Omar/kuberay anyscale

  #53318 commented on Jun 17, 2025 • 0 new comments

• [data] Add GroupedData.random_sample() for group-wise sampling

  #53313 commented on Jun 17, 2025 • 0 new comments

• [core] Core worker get cv - notify after unlock

  #53311 commented on Jun 17, 2025 • 0 new comments

• Make core worker testable

  #53299 commented on Jun 19, 2025 • 0 new comments

• [core][autoscaler][v1] drop object_store_memory from ResourceDemandScheduler._update_node_resources_from_runtime

  #53283 commented on Jun 18, 2025 • 0 new comments

• Bump tornado from 6.1 to 6.5.1 in /python

  #53274 commented on Jun 17, 2025 • 0 new comments

• Override Autoscaler

  #53245 commented on Jun 17, 2025 • 0 new comments

• [data] add explain interface for dataset

  #53235 commented on Jun 17, 2025 • 0 new comments

• [Serve] Support for setting `working_dir` to a local directory in `RayService`

  #33456 commented on Jun 17, 2025 • 0 new comments

• RLLIB - RE3 Exploration Algorithm - No GPU support f0r Dynamic TF V2

  #33425 commented on Jun 17, 2025 • 0 new comments

• [client] kubernetes w ray client

  #33367 commented on Jun 17, 2025 • 0 new comments

• [Train] Reporting metrics/checkpoints from multiple workers

  #33360 commented on Jun 17, 2025 • 0 new comments

• [Data] `read_parquet` schema is incorrect (schema is a dict instead of a string)

  #33279 commented on Jun 17, 2025 • 0 new comments

• [Ray status] confusing output about gpus and accelerators

  #33272 commented on Jun 17, 2025 • 0 new comments

• [Tune] mlflow logger callback > log_trial_result fail (psycopg2.ProgrammingError) can't adapt type 'numpy.int64'

  #33233 commented on Jun 17, 2025 • 0 new comments

• [Serve] Enhance replica upgrade process.

  #33192 commented on Jun 17, 2025 • 0 new comments

• [air output] Isolate/refactor/improve rllib related progress reporting logic

  #33150 commented on Jun 17, 2025 • 0 new comments

• [Tune][wandb] Report tune experiments as a wandb `sweep`

  #33142 commented on Jun 17, 2025 • 0 new comments

• [AIR][wandb] Add option to track artifact references in wandb if using cloud storage

  #33130 commented on Jun 17, 2025 • 0 new comments

• [AIR][Tune] Add an option in `WandbLoggerCallback` to group wandb runs by config

  #33084 commented on Jun 17, 2025 • 0 new comments

• [Serve] Support external storage for state

  #33059 commented on Jun 17, 2025 • 0 new comments

• [Serve] Use the namespace of context instead of "serve" when the Controller gets all running Actors

  #33057 commented on Jun 17, 2025 • 0 new comments

• [Serve] Specify replicas when scaling down

  #33056 commented on Jun 17, 2025 • 0 new comments

• [Serve] Restart a batch of replicas by Actor names or replica tags

  #33055 commented on Jun 17, 2025 • 0 new comments

• [Serve] Specify a batch of replicas to update their user_config

  #33054 commented on Jun 17, 2025 • 0 new comments

• Ray Core Runtime Environments with tea.xyz

  #33049 commented on Jun 17, 2025 • 0 new comments

• [Ray Tune] Support for continuing training when metrics are only reported from some of the workers

  #33042 commented on Jun 17, 2025 • 0 new comments

• [Data] Cannot get the length of a tf dataset created from `ray_ds.to_tf`

  #33004 commented on Jun 17, 2025 • 0 new comments

• [Data] Include image class id in the returned datasets of `ray.data.read_images()`.

  #32989 commented on Jun 17, 2025 • 0 new comments

• [Datasets] Raise descriptive error if `iter_torch_batches` can't convert data

  #32953 commented on Jun 17, 2025 • 0 new comments

• [RLlib] Windows CLI, cmd.exe, powershell parsing json arguments JSONDecodeError

  #35492 commented on Jun 17, 2025 • 0 new comments

• [Core] Timeout for unschedulable task due to unavailable workers

  #33954 commented on Jun 17, 2025 • 0 new comments

• [Observability] Programmatically fetch prometheus metrics

  #33940 commented on Jun 17, 2025 • 0 new comments

• [Ray AIR] Add more documentation about checkpointing

  #33932 commented on Jun 17, 2025 • 0 new comments

• Ray Workflow

  #33844 commented on Jun 17, 2025 • 0 new comments

• [Train] Intermittent `UnpicklingError` when loading estimator/preprocessor from checkpoint

  #33815 commented on Jun 17, 2025 • 0 new comments

• [AIR output] Warnings for AIR_VERBOSITY is confusing

  #33810 commented on Jun 17, 2025 • 0 new comments

• [air output] Aggregation of feedback for air output v2

  #33803 commented on Jun 17, 2025 • 0 new comments

• [Datasets] `FileBasedDataSource`s do not pass `filesystem` to `_read_stream()` methods' `reader_args`

  #33777 commented on Jun 17, 2025 • 0 new comments

• [Core][Runtime Env] Document how to write custom runtime env plugin

  #33746 commented on Jun 17, 2025 • 0 new comments

• Core: Can the ray core's scheduling mechanism support customized extensions?

  #33735 commented on Jun 17, 2025 • 0 new comments

• [Ray init] Ray init method does not support pathlib.Path

  #33672 commented on Jun 17, 2025 • 0 new comments

• [docs] improve user experience of the API ref

  #33645 commented on Jun 17, 2025 • 0 new comments

• [RLLib] Collecting external experience

  #33636 commented on Jun 17, 2025 • 0 new comments

• [Workflow] get_metadata(workflow_id)["status"] and get_status(workflow_id) not returning the same status

  #33633 commented on Jun 17, 2025 • 0 new comments

• [runtime_env] Actors always depend global `pip` field for `runtime_env`

  #33607 commented on Jun 17, 2025 • 0 new comments

• [Core] Raylet process not respecting `--node-ip-address`

  #33554 commented on Jun 17, 2025 • 0 new comments

• [Tune] Support ExperimentAnalysis.dataframe(mode='mean')

  #33540 commented on Jun 17, 2025 • 0 new comments

• [Train] `RunConfig` doesn't get propagated from the Tuner to the Trainer

  #33539 commented on Jun 17, 2025 • 0 new comments

• [Core] std::bad_alloc error using ray.init()

  #33525 commented on Jun 17, 2025 • 0 new comments

• [Core] `test_memory_deadlock` times out

  #33491 commented on Jun 17, 2025 • 0 new comments

• [Core] Support binding worker processes to NUMA nodes

  #33465 commented on Jun 17, 2025 • 0 new comments

• Serve build usage of click CLI library conflicts python argparse

  #32001 commented on Jun 17, 2025 • 0 new comments

• [Serve] Version Support in 2.X API

  #31928 commented on Jun 17, 2025 • 0 new comments

• [Train] User exceptions not propagated from remote cluster

  #31913 commented on Jun 17, 2025 • 0 new comments

• [RLlib] AlgorithmConfig() defaults not used by build_sac_model when implementing custom model

  #31783 commented on Jun 17, 2025 • 0 new comments

• [kubernetes/cluster] More guides on deployment

  #31623 commented on Jun 17, 2025 • 0 new comments

• [core][state] ray log supporting regex searching

  #31549 commented on Jun 17, 2025 • 0 new comments

• [Tune] Support NLopt search algorithms

  #31492 commented on Jun 17, 2025 • 0 new comments

• [Rllib] Possible Redudant Code

  #31463 commented on Jun 17, 2025 • 0 new comments

• [aws] ray submit --stop fails on aws

  #31380 commented on Jun 17, 2025 • 0 new comments

• [Tune] Avoid insufficient resources warning if cluster is autoscaling

  #31292 commented on Jun 17, 2025 • 0 new comments

• No worker logs in the dashboard after recreating the K8S Ray pods

  #31288 commented on Jun 17, 2025 • 0 new comments

• [core] Please improve warning message for ip mismatch

  #31264 commented on Jun 17, 2025 • 0 new comments

• [Ray Collective] Ray Collective AllGather is Completely Broken

  #31259 commented on Jun 17, 2025 • 0 new comments

• [core][state] Refactor use of bounded LRU/FIFO buffer/map used in task backend

  #31158 commented on Jun 17, 2025 • 0 new comments

• [core] Ray resources should be case-insensitive

  #31087 commented on Jun 17, 2025 • 0 new comments

• [RayCluster]

  #31041 commented on Jun 17, 2025 • 0 new comments

• [Serve] gRPCis should not allow route_prefix set

  #30891 commented on Jun 17, 2025 • 0 new comments

• [General] Setup a "code walkthrough" meetup or tutorial

  #30852 commented on Jun 17, 2025 • 0 new comments

• [RFC][core] Option to avoid scheduling tasks to nodes with disk full

  #30843 commented on Jun 17, 2025 • 0 new comments

• [core] Enable greater control over log verbosity

  #30832 commented on Jun 17, 2025 • 0 new comments

• [Core] ux issues of ray state cli for tasks

  #30805 commented on Jun 17, 2025 • 0 new comments

• [Tune] ability to specify search algorithm when using tune.run_experiments()

  #30802 commented on Jun 17, 2025 • 0 new comments

• [Serve] Don't start Serve agent if Serve isn't installed

  #32920 commented on Jun 17, 2025 • 0 new comments

• [Data]: `ds.take()` and `ds.iter_batches()` have unexpected different behavior for pd.Series columns

  #32913 commented on Jun 17, 2025 • 0 new comments

• [Ray: Serve] Model Composition primitives should be part of Serve Core API docs.

  #32837 commented on Jun 17, 2025 • 0 new comments

• [core][state] Task backend : already submitted cancelled task showing up as finished

  #32826 commented on Jun 17, 2025 • 0 new comments

• [AIR][Tune] Make trial checkpoint + artifact upload happen atomically

  #32823 commented on Jun 17, 2025 • 0 new comments

• [Tune] During multi-GPU training (using mp.spawn), ray.tune.report does not take effect.

  #32810 commented on Jun 17, 2025 • 0 new comments

• [Tune] failure when using more than one GPU

  #32760 commented on Jun 17, 2025 • 0 new comments

• [Runtime Env] Add docstring for public class methods and attributes

  #32704 commented on Jun 17, 2025 • 0 new comments

• [tune] Add suggestions on when `reuse_actor` should be set to false.

  #32698 commented on Jun 17, 2025 • 0 new comments

• [serve] serve run doesn't restart app successfully in some environments

  #32633 commented on Jun 17, 2025 • 0 new comments

• [train] Big performance hit when TensorFlow trainer is not scheduled on head node

  #32509 commented on Jun 17, 2025 • 0 new comments

• [doc][tune] clarify `Stopper`, what is `training_iteration`

  #32497 commented on Jun 17, 2025 • 0 new comments

• [release] update our xgboost release test to catch issues like (see discription)

  #32491 commented on Jun 17, 2025 • 0 new comments

• [Core] The remote function in the worker no longer runs after the head crashes

  #32454 commented on Jun 17, 2025 • 0 new comments

• [RLlib] Special __common__ key in MultiAgent batches is not documented

  #32399 commented on Jun 17, 2025 • 0 new comments

• [tune] update how trainable reports result/checkpoint to driver

  #32380 commented on Jun 17, 2025 • 0 new comments

• [Datasets] The projection pushdown cannot work with hive style partitioning file path

  #32301 commented on Jun 17, 2025 • 0 new comments

• [Core][utilization] some anti-pattern that not well supported by Ray core.

  #32297 commented on Jun 17, 2025 • 0 new comments

• [tune/train] Provide actionable error messages for common thirdparty errors

  #32232 commented on Jun 17, 2025 • 0 new comments

• [ci] Mirror external dependenies in CI

  #32113 commented on Jun 17, 2025 • 0 new comments

• [Serve] ValueError: Message ray.serve.ReplicaConfig exceeds maximum protobuf size of 2GB

  #32049 commented on Jun 17, 2025 • 0 new comments

• [CLI] make `ray get-head-ip` and `ray get-worker-ips` work for kuberay clusters when run outside the cluster

  #32037 commented on Jun 17, 2025 • 0 new comments

• [RayClient]large object transfer failure

  #35448 commented on Jun 17, 2025 • 0 new comments

• [train] Simplify `test_transformers_trainer_steps::test_e2e_steps`

  #35424 commented on Jun 17, 2025 • 0 new comments

• [Core] Reducing scheduling fragmentation

  #35422 commented on Jun 17, 2025 • 0 new comments

• [Core, RLlib] Multi GPU RLlib experiment is unable to be scheduled.

  #35409 commented on Jun 17, 2025 • 0 new comments

• [Job] Failed to schedule supervisor actor leads to job failure

  #35387 commented on Jun 17, 2025 • 0 new comments

• [Job] Show submitter of a Job on the dashboard

  #35367 commented on Jun 17, 2025 • 0 new comments

• [Serve] Support sync function for multiplexing

  #35356 commented on Jun 17, 2025 • 0 new comments

• [AIR] [Train] train multiple instances simultaneously on machines with specified tags

  #35333 commented on Jun 17, 2025 • 0 new comments

• <RLlib> What is the cause of the low CPU utilization in rllib PPO?

  #35313 commented on Jun 17, 2025 • 0 new comments

• [VM launcher] Document how to set up the cluster when there is UFW firewall

  #35254 commented on Jun 17, 2025 • 0 new comments

• [Data] Infer the data schema in Ray Datasets

  #35230 commented on Jun 17, 2025 • 0 new comments

• [dashboard] how to adjust ray dashboard refresh rate?

  #35156 commented on Jun 17, 2025 • 0 new comments

• [KubeRay, dashboard] Clarify that the users can use persistent volumes for log_dir and ray dashboard can read from it.

  #35137 commented on Jun 17, 2025 • 0 new comments

• [RLlib] Better error handling when return shape from step() mismatch in utils._flatten_multidiscrete

  #35113 commented on Jun 17, 2025 • 0 new comments

• The ray rsync-up cli reports no issue, but actually file is absent on remote side (Ray AWS cluster)

  #35051 commented on Jun 17, 2025 • 0 new comments

• [Core] - GPU Support - Explanation of Results

  #35048 commented on Jun 17, 2025 • 0 new comments

• [Data] Optimize `read_datasource` setup

  #35029 commented on Jun 17, 2025 • 0 new comments

• [EC2 VM Cluster launcher] Document EC2 ssh key limit and workaround

  #35020 commented on Jun 17, 2025 • 0 new comments

• [VM launcher] Ran `Ray status` after I sshed in to the head node and it printed "No cluster status"

  #35017 commented on Jun 17, 2025 • 0 new comments

• [air/tune][multi-tenancy] Parallel runs can use the same experiment directory

  #35006 commented on Jun 17, 2025 • 0 new comments

• Issue on page /cluster/vms/examples/ml-example.html

  #34996 commented on Jun 17, 2025 • 0 new comments

• [Core] Custom docker image not scaling out

  #53696 commented on Jun 17, 2025 • 0 new comments

• [Serve] `fastapi_app` is still mutable in the deployment constructor after being passed to `@serve.ingress`

  #52775 commented on Jun 17, 2025 • 0 new comments

• [tune] `URI has empty scheme` error when `storage_path` in `RunConfig` is relative

  #42969 commented on Jun 17, 2025 • 0 new comments

• [Serve] Specify different images for each deployment

  #52994 commented on Jun 17, 2025 • 0 new comments

• [Conda] Ray should raise exception when ray is not installed in conda environment

  #52672 commented on Jun 17, 2025 • 0 new comments

• [Serve] Autoscaling not working correctly when `max_replica_per_node` is set in Ray Serve

  #53582 commented on Jun 17, 2025 • 0 new comments

• [Serve] Allow --metrics-export-port argument in "serve run" CLI command

  #44426 commented on Jun 17, 2025 • 0 new comments

• [data] verbose_progress=True doesn't work in client mode

  #43200 commented on Jun 17, 2025 • 0 new comments

• [data] importing ray.data closes logging handlers, breaking custom logging

  #48846 commented on Jun 17, 2025 • 0 new comments

• Ray Serve Replica Initialization Timeout: STDOUT "Failed to load", RequestCancelledError, Likely Due to Slow/Crashing RLModule.from_checkpoint()

  #53079 commented on Jun 17, 2025 • 0 new comments

• [RLlib] TorchDistributionWrapper Typing Information Should Be Changed

  #33997 commented on Jun 17, 2025 • 0 new comments

• [Core] DecodeError when `ray.put` a large (2GB) object

  #35976 commented on Jun 17, 2025 • 0 new comments

• [core][gpu-objects] Support streaming to overlap computation / communication

  #51643 commented on Jun 17, 2025 • 0 new comments

• [core][gpu-objects] Allow tensor metadata to be specified ahead of time for improved performance

  #51279 commented on Jun 17, 2025 • 0 new comments

• [<Ray component: Core|RLlib|etc...>] SAC config error about framework

  #53694 commented on Jun 17, 2025 • 0 new comments

• [air/output] Jupyter notebook trial result table keeps swapping column order

  #35838 commented on Jun 17, 2025 • 0 new comments

• [RLlib] Make Learner more standalone with regards to LearnerHyperparameters

  #35788 commented on Jun 17, 2025 • 0 new comments

• [AIR] `on_trial_complete` callback hook happens before trial resources are freed

  #35721 commented on Jun 17, 2025 • 0 new comments

• [core] Failed to close sockets in CoreWorker when crash.

  #35681 commented on Jun 17, 2025 • 0 new comments

• [serve][dashboard] Show last line instead of first line in Serve app status message

  #35600 commented on Jun 17, 2025 • 0 new comments

• Ray Data - Glob/wildcard in file path

  #35499 commented on Jun 17, 2025 • 0 new comments

• [serve] Document how to silence access logs from GradioIngress

  #35496 commented on Jun 17, 2025 • 0 new comments

• [Serve] Production Guide: Add instruction for non-K8s on-premise clusters

  #34437 commented on Jun 17, 2025 • 0 new comments

• [Serve] Ray Serve hangs and becomes unresponsive when calling ffmpeg in deployment

  #34414 commented on Jun 17, 2025 • 0 new comments

• [Serve] Deployments page tasks history is full of system tasks. Not very useful

  #34386 commented on Jun 17, 2025 • 0 new comments

• [Core] serialisation of dataclass in separate module fails to recognise parameter change in child dataclass, but functions correctly if in the same module

  #34366 commented on Jun 17, 2025 • 0 new comments

• ImportError: cannot import name 'torch' from 'ray.rllib.train'

  #34354 commented on Jun 17, 2025 • 0 new comments

• [core][state] Include job info for placement group

  #34333 commented on Jun 17, 2025 • 0 new comments

• [Jobs] Use new API `is_head_node` to find head node

  #34317 commented on Jun 17, 2025 • 0 new comments

• [Core] RFC: simplify CI testing

  #34315 commented on Jun 17, 2025 • 0 new comments

• [air] Error while loading xgboost model in BatchPredictor

  #34307 commented on Jun 17, 2025 • 0 new comments

• [RLlib] Unity 3d env tests are broken

  #34290 commented on Jun 17, 2025 • 0 new comments

• [air/train] the logic to grab free ports for `tf_config` is potentially racy

  #34271 commented on Jun 17, 2025 • 0 new comments

• [Core][Object Store] Push Manager: round for object manager client and FIFO for object

  #34270 commented on Jun 17, 2025 • 0 new comments

• [air] xgboost/lightgbm trainer's validation result differ between online and offline

  #34211 commented on Jun 17, 2025 • 0 new comments

• [tune] support viewing partial experiment result as tuning goes on

  #34207 commented on Jun 17, 2025 • 0 new comments

• [Workflow] Improve efficiency of Ray Workflow by returning workflow metadata and completed task information in single API call

  #34158 commented on Jun 17, 2025 • 0 new comments

• Issue on page /rllib/package_ref/algorithm.html

  #34157 commented on Jun 17, 2025 • 0 new comments

• [Prometheus metrics util] Application level custom metrics aren't getting exported consistently

  #34145 commented on Jun 17, 2025 • 0 new comments

• [Core] Actors not cleaning up resources correct because `force_kill=true`.

  #34124 commented on Jun 17, 2025 • 0 new comments

• Ray Tune + ray xgboost running out of disk space

  #34118 commented on Jun 17, 2025 • 0 new comments

• [Core][Tune]Trials hang when using Pytorch

  #34028 commented on Jun 17, 2025 • 0 new comments

• [Data] `map_batches` hard to use and debug

  #34007 commented on Jun 17, 2025 • 0 new comments

• [Core] improve garbage collection after job go out of scope

  #34001 commented on Jun 17, 2025 • 0 new comments

• [AWS VM Cluster Launcher] AWS Cluster launcher installs nightly Ray by default

  #34991 commented on Jun 17, 2025 • 0 new comments

• [CI] Fix minimal-install python 3.11: build wheel with unsupported tags.

  #34980 commented on Jun 17, 2025 • 0 new comments

• [serve][docs] Add DAG building classes to the API reference

  #34953 commented on Jun 17, 2025 • 0 new comments

• [AIR output] Rich table gets truncated when the terminal height is smaller than it

  #34925 commented on Jun 17, 2025 • 0 new comments

• [AIR output] Format of trial table with Rich enabled.

  #34923 commented on Jun 17, 2025 • 0 new comments

• [AIR output] "iteration" is shown in the output for RL users

  #34918 commented on Jun 17, 2025 • 0 new comments

• [core] ray.kill doesn't guarantee resources are cleaned up

  #34917 commented on Jun 17, 2025 • 0 new comments

• [Data] Add `fn_kwargs` to `BatchMapper`

  #34852 commented on Jun 17, 2025 • 0 new comments

• Resource Allocation: Ray Core, Ray Client

  #34816 commented on Jun 17, 2025 • 0 new comments

• [Jobs] Job agent recovers all running jobs on restart, not just those monitored by that agent

  #34794 commented on Jun 17, 2025 • 0 new comments

• [Doc] Autogenerated "suggest an edit" link doesn't work

  #34751 commented on Jun 17, 2025 • 0 new comments

• [Tune] thread limit resulting in the job failure in multi-tenancy usage

  #34745 commented on Jun 17, 2025 • 0 new comments

• Ray Job

  #34710 commented on Jun 17, 2025 • 0 new comments

• [docs][infra] automate checks for common link errors

  #34681 commented on Jun 17, 2025 • 0 new comments

• [Ray Job] Auto-shutdown of the cluster when job finished

  #34672 commented on Jun 17, 2025 • 0 new comments

• [Core] Ray.wait should return if task throw exception

  #34653 commented on Jun 17, 2025 • 0 new comments

• [Core] ray2.3.1 gcs_server memory keeps increasing until OOM

  #34619 commented on Jun 17, 2025 • 0 new comments

• [Runtime Env/Ray Job] Job submission fails when specifing local zip file as working dir

  #34605 commented on Jun 17, 2025 • 0 new comments

• why ray.data.read_images cat not combine_chunks

  #34563 commented on Jun 17, 2025 • 0 new comments

• [Core] Add support for cancelling descendants of a completed task

  #34545 commented on Jun 17, 2025 • 0 new comments

• [Data] retrieve written paths from `Dataset.write_datasource`

  #34444 commented on Jun 17, 2025 • 0 new comments

• [Docs Infra] [RLLib] Remove "<<<" from code blocks

  #34439 commented on Jun 17, 2025 • 0 new comments

• [docs][Bug] Workflow docs have few typos and type issue

  #23113 commented on Jun 16, 2025 • 0 new comments

• [tune][Feature] add tune.choices to select multiple values from a search space

  #23001 commented on Jun 16, 2025 • 0 new comments

• [Bug] An exception in a task cannot be caught with ActorPool.map_unordered making restarting meaningless

  #22978 commented on Jun 16, 2025 • 0 new comments

• Ray Train / Tune - W&B logger documentation

  #22881 commented on Jun 16, 2025 • 0 new comments

• [Train] update `logdir` relative path

  #22753 commented on Jun 16, 2025 • 0 new comments

• [Bug][placement groups] Actor scheduling does not respect placement_group=None

  #22742 commented on Jun 16, 2025 • 0 new comments

• [Cluster snapshot] [Bug] `runtime_env` fields in cluster snapshot are converted to camelcase when they should not be

  #22565 commented on Jun 16, 2025 • 0 new comments

• [Train] Add flags to disable creating log directories

  #22261 commented on Jun 16, 2025 • 0 new comments

• [RLlib] [Feature] Support for having parametric action spaces/action masking for continuous action space models

  #22259 commented on Jun 16, 2025 • 0 new comments

• [RLLib] Workers died at the initialization stage when the observation space is a 3D shape

  #22033 commented on Jun 16, 2025 • 0 new comments

• [Train] Automatically choose number of workers

  #21987 commented on Jun 16, 2025 • 0 new comments

• [Serve] The adjustment about Ray Serve Java Proxy and Java Replica

  #21694 commented on Jun 16, 2025 • 0 new comments

• [C++] Cluster Mode Tests Should have 1 test per feature tested

  #21454 commented on Jun 16, 2025 • 0 new comments

• [Tune] [Bug] lazily expand directories for client compatibility

  #21408 commented on Jun 16, 2025 • 0 new comments

• [Tune] Issue on page /tune/tutorials/tune-pytorch-lightning.html

  #21354 commented on Jun 16, 2025 • 0 new comments

• [Bug] Got stucked when running python script from a shell script

  #21298 commented on Jun 16, 2025 • 0 new comments

• [Bug] [Tune] pbt run_experiments not stable, some trial will error.

  #21259 commented on Jun 16, 2025 • 0 new comments

• [Train] Document Callbacks

  #21066 commented on Jun 16, 2025 • 0 new comments

• [Feature] Single source of truth for Ray version in Java `pom.xml` and `pom_template.xml` files

  #21059 commented on Jun 16, 2025 • 0 new comments

• [Test Bug] Matching `psutil.Process.name()` doesn't work on macOS

  #20982 commented on Jun 16, 2025 • 0 new comments

• [Bug] Incorrect promise usage that causes infinite blocking calls

  #20899 commented on Jun 16, 2025 • 0 new comments

• We encountered the cast exception after we got result from ray actor task

  #20369 commented on Jun 16, 2025 • 0 new comments

• [Train/AIR] Ray Train actors still use up resources after Notebook cell is stopped

  #24947 commented on Jun 16, 2025 • 0 new comments

• [Core] Failed to delete named actor in client mode

  #24906 commented on Jun 16, 2025 • 0 new comments

• [AIR] Add a `reconfigure` option to `ModelWrapperDeployment`

  #24869 commented on Jun 16, 2025 • 0 new comments

• [core] Uninformative error for unserialisable objects

  #24863 commented on Jun 16, 2025 • 0 new comments

• [Serve] Prototype C++ Worker in Serve

  #24738 commented on Jun 16, 2025 • 0 new comments

• [Core] Spilling performance regression in large-scale shuffle

  #24667 commented on Jun 16, 2025 • 0 new comments

• [Core] Restore objects directly from S3

  #24581 commented on Jun 16, 2025 • 0 new comments

• [Ray component: Core] Dask on Ray - Worker processes go to idle state and not garbage collected when used with RayProgressBar()

  #24556 commented on Jun 16, 2025 • 0 new comments

• [runtime env] `serialized_env` used as ID, but identical envs can produce different `serialized_env`

  #24515 commented on Jun 16, 2025 • 0 new comments

• [Core] No overloads for "remote" match the provided arguments

  #24371 commented on Jun 16, 2025 • 0 new comments

• Workflows: Type stubs are incorrect: argument missing for parameter status_filter

  #24367 commented on Jun 16, 2025 • 0 new comments

• [Core] /api/cluster_status treats placement groups differently than ray status

  #24309 commented on Jun 16, 2025 • 0 new comments

• [Core] Restore worker silently fails and the program is stuck

  #24248 commented on Jun 16, 2025 • 0 new comments

• [RLlib][Bug] duplicate action unsquashing in DDPG / TD3 policy

  #24213 commented on Jun 16, 2025 • 0 new comments

• [Tune] support for FIRE PBT

  #24137 commented on Jun 16, 2025 • 0 new comments

• [Tune] Tune Job hangs out and can't finish the tune job

  #23858 commented on Jun 16, 2025 • 0 new comments

• [Workflows] Cant use custom storage backends

  #23831 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Add Option for Custom Sample Preprocessing when Sampling from Replay Buffer

  #23815 commented on Jun 16, 2025 • 0 new comments

• [Core][Bug] global-scoped actor handles/Ray objects prevents Ray workers from being destructed.

  #23677 commented on Jun 16, 2025 • 0 new comments

• [runtime env] `zip_directory` `excludes` parameter doesn't work with absolute paths

  #23473 commented on Jun 16, 2025 • 0 new comments

• [Train] [Feature] Print useful traceback on SIGINT

  #23148 commented on Jun 16, 2025 • 0 new comments

• [Train] [Docs] Document how to change logging verbosity

  #23147 commented on Jun 16, 2025 • 0 new comments

• [Train] Refactor `TrainingIterator` result processing logic

  #20330 commented on Jun 16, 2025 • 0 new comments

• [serve] java api

  #16393 commented on Jun 16, 2025 • 0 new comments

• [serve] java serve handle

  #16392 commented on Jun 16, 2025 • 0 new comments

• [serve] java http proxy

  #16391 commented on Jun 16, 2025 • 0 new comments

• [Shuffle] non-streaming consumed bytes are too low compared to spilled / restored bytes.

  #16149 commented on Jun 16, 2025 • 0 new comments

• [ray] Multiple concurrent requests to create a named actor crash GCS

  #15941 commented on Jun 16, 2025 • 0 new comments

• Remove unused util functions for conda environments

  #15912 commented on Jun 16, 2025 • 0 new comments

• [core] Zero-gpu node shouldn't be marked with accelerator_type resource.

  #15878 commented on Jun 16, 2025 • 0 new comments

• Cannot using external model with cuda when using ray

  #15869 commented on Jun 16, 2025 • 0 new comments

• [wheel][doc] Make it easier to access Ray wheels for specific commits

  #15765 commented on Jun 16, 2025 • 0 new comments

• [rllib]Update the docs about Variable-length / Parametric Action Space

  #15710 commented on Jun 16, 2025 • 0 new comments

• Odd task scheduling behavior on same node

  #15602 commented on Jun 16, 2025 • 0 new comments

• Averaging learning curves over repetitions + plotting confidence intervals [Tune]

  #15400 commented on Jun 16, 2025 • 0 new comments

• AssertionError when using pyinstaller with ray

  #15396 commented on Jun 16, 2025 • 0 new comments

• [core] Memory leak when using local simulated cluster (long_running_tests/workloads/apex.py)

  #15305 commented on Jun 16, 2025 • 0 new comments

• [Core] Bad traceback on failure to reconnect to GCS server.

  #15235 commented on Jun 16, 2025 • 0 new comments

• [metrics] Custom sum metrics have type comment "gauge"

  #15150 commented on Jun 16, 2025 • 0 new comments

• [core] Actor restart does not work when owner dies and constructor task has dependencies

  #15076 commented on Jun 16, 2025 • 0 new comments

• [k8s] ray down command does not remove pods which are in evicted state

  #14958 commented on Jun 16, 2025 • 0 new comments

• [Tune] [Ray Client] tune_cifar10_gluon example fails with Ray Client

  #14946 commented on Jun 16, 2025 • 0 new comments

• [ray white paper] broken links

  #14897 commented on Jun 16, 2025 • 0 new comments

• Fix Asyncio Event Metrics on Java

  #14715 commented on Jun 16, 2025 • 0 new comments

• Add ray.__wheel__ with a link to the wheel to install the same version

  #14623 commented on Jun 16, 2025 • 0 new comments

• [tsan] Add TSAN CI build that runs basic Python tests

  #20080 commented on Jun 16, 2025 • 0 new comments

• [tsan] Race in census SetGlobalTags

  #20079 commented on Jun 16, 2025 • 0 new comments

• [tsan] Race accessing global stats objects

  #20078 commented on Jun 16, 2025 • 0 new comments

• [tsan] Several global config variables accessed unsafely

  #20077 commented on Jun 16, 2025 • 0 new comments

• Support working_dir=None for skipping packaging upload/download

  #19962 commented on Jun 16, 2025 • 0 new comments

• [Bug] Placement group removal refinement

  #19937 commented on Jun 16, 2025 • 0 new comments

• [Feature] Able to access objects put in cross language

  #19873 commented on Jun 16, 2025 • 0 new comments

• [Bug] Improve RuntimeEnvSetupError message

  #19824 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Deprecate Internally Maintained Probability Distributions In Favor Of Native TFP And torch.distributions Solutions

  #19725 commented on Jun 16, 2025 • 0 new comments

• [Serve] Test KVStore early in constructor init.

  #19714 commented on Jun 16, 2025 • 0 new comments

• [Bug] [Workflow] ray.wait on workflow result doesn't work as expected

  #19295 commented on Jun 16, 2025 • 0 new comments

• [tune] MLFlowLogger doesn't save artifacts for remote mlflow tracking_uri

  #19263 commented on Jun 16, 2025 • 0 new comments

• [Bug] [XLang] Segfault when Java returns void

  #18837 commented on Jun 16, 2025 • 0 new comments

• [Bug] tensorboardX vs tensorboard?

  #18727 commented on Jun 16, 2025 • 0 new comments

• Dashboard exposes redis PW on the command line

  #18491 commented on Jun 16, 2025 • 0 new comments

• [core] ReferenceCountingAssertionError may be thrown if ObjectRef is passed through intermediate worker that dies

  #18456 commented on Jun 16, 2025 • 0 new comments

• Race condition of grpc backpressure

  #18439 commented on Jun 16, 2025 • 0 new comments

• [Core] Task spec including inlined objects can crash lease request RPCs.

  #18194 commented on Jun 16, 2025 • 0 new comments

• [Runtime Env] Setup process doesn't have CPU limit

  #18137 commented on Jun 16, 2025 • 0 new comments

• ray.init with address crashes process outside of cluster

  #17769 commented on Jun 16, 2025 • 0 new comments

• new dashboard agent port conflict issues

  #17498 commented on Jun 16, 2025 • 0 new comments

• [Core] Unable to get actor handle of global named actor created in java from python in Ray 1.4.0

  #16436 commented on Jun 16, 2025 • 0 new comments

• [Clusters] [KubeRay] problem with pending actors' pods in Kubernetes

  #32651 commented on Jun 16, 2025 • 0 new comments

• [core] Lock contention when submitting actor task on the client queue

  #32595 commented on Jun 16, 2025 • 0 new comments

• [Core] Install via `pip` fails, install with `conda` crashes worker and exits

  #32423 commented on Jun 16, 2025 • 0 new comments

• [Core] "ImportError: No module named ray" when using `ray submit`

  #31924 commented on Jun 16, 2025 • 0 new comments

• [workflow] memory leakage

  #31819 commented on Jun 16, 2025 • 0 new comments

• [Clusters] [RLlib] Trainer Object running on Worker node & RolloutWorker running on Head node

  #31808 commented on Jun 16, 2025 • 0 new comments

• [runtime envs] Ray Client Server failed when starting

  #31622 commented on Jun 16, 2025 • 0 new comments

• Huge numbers of "deleted" files with open processes left after Ray Tune run

  #31556 commented on Jun 16, 2025 • 0 new comments

• [Tune] Reenable `zoopt` searcher test after fixes for handling invalid results are included in its next release

  #31439 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Pytorch multiple optimizers

  #31428 commented on Jun 16, 2025 • 0 new comments

• In the docker bridge mode, pulling the actor on a non head node fails.

  #31308 commented on Jun 16, 2025 • 0 new comments

• [CORE] Unable to run celery task containing ray tasks

  #31157 commented on Jun 16, 2025 • 0 new comments

• [core] Segfaults when restarting Ray multiple times in unit tests with background threads running

  #31145 commented on Jun 16, 2025 • 0 new comments

• [core] Error with Slurm: No available node types can fulfill resource request {'node:<ip>': 0.01}.

  #31135 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Not able to save evaluation recording videos

  #30949 commented on Jun 16, 2025 • 0 new comments

• [Ray Job] SchedulingCancelled for JobSupervisor Actor

  #30898 commented on Jun 16, 2025 • 0 new comments

• [Ray client] Ray Zombie Process Issue

  #30894 commented on Jun 16, 2025 • 0 new comments

• [Devprod] Bazel reports an error when compiling as a non-root user

  #30885 commented on Jun 16, 2025 • 0 new comments

• [release tests] Prometheus metrics collection sometimes takes 15min to run for long_running_node_failures

  #30859 commented on Jun 16, 2025 • 0 new comments

• [core] Disk full error logging is verbose

  #30833 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Error when running RLlib

  #30412 commented on Jun 16, 2025 • 0 new comments

• [Cluster Launcher] `ray dashboard` CLI command does not stop port-forwarding after Ctrl+C

  #30385 commented on Jun 16, 2025 • 0 new comments

• Ray: Data - Cannot read json its written to s3

  #35501 commented on Jun 16, 2025 • 0 new comments

• [Core] `OwnerDiedError` if dataset owner actor handle get out of scope

  #35262 commented on Jun 16, 2025 • 0 new comments

• [VM launcher] Automtically shut down the ec2 machine when I stop ray up in the middle

  #35013 commented on Jun 16, 2025 • 0 new comments

• [CI] Migrate from flake8 to ruff

  #34889 commented on Jun 16, 2025 • 0 new comments

• [Core] Incorrect detection of cpus

  #34846 commented on Jun 16, 2025 • 0 new comments

• [Clusters] - Cannot switch off rsync during Cluster Launch with `ray up`

  #34390 commented on Jun 16, 2025 • 0 new comments

• Azure autoscaler cannot create additional nodes

  #34198 commented on Jun 16, 2025 • 0 new comments

• [Core] Error in external storage writing for object spilling

  #33913 commented on Jun 16, 2025 • 0 new comments

• get_node_to_storage_syncer has an empty docstring

  #33841 commented on Jun 16, 2025 • 0 new comments

• [ Core ] Correct usage of min/max-worker-port arguments

  #33749 commented on Jun 16, 2025 • 0 new comments

• Core: nightly builds for macos only include an x86 _raylet.so even though they claim to be universal

  #33720 commented on Jun 16, 2025 • 0 new comments

• [Core] The resources have minus values in ray status output

  #33569 commented on Jun 16, 2025 • 0 new comments

• [tune] tqdm/Hyperopt-style TuneReporter for Databricks notebooks

  #33519 commented on Jun 16, 2025 • 0 new comments

• [Core] Ray client doesn't support `should_capture_child_tasks_in_placement_group` API

  #33513 commented on Jun 16, 2025 • 0 new comments

• [RLlib] DictFlatteningPreprocessor order is inconsistent leads to invalid mapping of OBS

  #33327 commented on Jun 16, 2025 • 0 new comments

• [Dashboard] Head node exited unexceptly because of dashboard process exited

  #31261 commented on Jun 16, 2025 • 0 new comments

• [runtime env] Raise warning when using `runtime_env` with `local_mode=True`

  #33260 commented on Jun 16, 2025 • 0 new comments

• [Core] `get_runtime_context()` in task fails with unhelpful error "cannot pickle '_thread.lock' object"

  #32987 commented on Jun 16, 2025 • 0 new comments

• [Train] Benchmark testing on Mosaic Composer with Ray

  #32946 commented on Jun 16, 2025 • 0 new comments

• [Ray Core] Actor Handles not properly passed to Actors created by other Actors

  #32848 commented on Jun 16, 2025 • 0 new comments

• [RLlib] A3C has problems with the horizon option removed

  #32812 commented on Jun 16, 2025 • 0 new comments

• [Core][Object Store] Object Store to manage files in the cluster

  #32694 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] AWS Single Sign-On support

  #30064 commented on Jun 16, 2025 • 0 new comments

• [Ray: Core] - Unable to enable TLS on the ray head node

  #28534 commented on Jun 16, 2025 • 0 new comments

• Dashboard / Jobs RegexMatcher ignores "includes".

  #28502 commented on Jun 16, 2025 • 0 new comments

• [Core, RLlib] RLlib uses Metal GPU even when told not to

  #28385 commented on Jun 16, 2025 • 0 new comments

• [Core] Actor methods will be modified for tracing even if tracing is not enabled.

  #28293 commented on Jun 16, 2025 • 0 new comments

• [Runtime] Improve runtime environment error message when virtualenv version is too old

  #28232 commented on Jun 16, 2025 • 0 new comments

• [Core] Multi-Threaded Actors are Un-Killable

  #28086 commented on Jun 16, 2025 • 0 new comments

• [Autoscaler] Assigning None to optional keys leads to failure

  #28012 commented on Jun 16, 2025 • 0 new comments

• [Core] Can't pickle objects defined in top-level environment

  #28000 commented on Jun 16, 2025 • 0 new comments

• [Doc] [Serve] Serve Loki monitoring tutorial screenshot has outdated API

  #27453 commented on Jun 16, 2025 • 0 new comments

• [core] Very slow task scheduling during Dataset.sort on 100TB

  #27410 commented on Jun 16, 2025 • 0 new comments

• Is Ray going to support Weighted Quantile Sketches or Quantile Sketches?

  #27363 commented on Jun 16, 2025 • 0 new comments

• [Core] Raylet continually exiting on worker in docker

  #26576 commented on Jun 16, 2025 • 0 new comments

• Tensorboard with Docker from Ray dashboard, tune tab cannot be accessed

  #26325 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Eval episode runs forever if Env doesn't terminate properly

  #26241 commented on Jun 16, 2025 • 0 new comments

• [Ray Client] Using many concurrent client connections results in deadlock/hanging

  #26144 commented on Jun 16, 2025 • 0 new comments

• [Core] worker died randomly and unexpectedly under heavy workload (Check failed: inner_it->second.mutable_nested()->contained_in_borrowed_ids.erase(id))

  #26128 commented on Jun 16, 2025 • 0 new comments

• [Core][HA] Actor entries are not deleted from the storage permanently if GCS is crashed.

  #26114 commented on Jun 16, 2025 • 0 new comments

• Unclear error when using generator tasks

  #25836 commented on Jun 16, 2025 • 0 new comments

• [Core] SIGSEGV when I run experimental shuffle command.

  #25650 commented on Jun 16, 2025 • 0 new comments

• [Core][Metrics] Prometheus-client not working with the latest version.

  #25523 commented on Jun 16, 2025 • 0 new comments

• [core] Scheduler stalls during shuffle reduce stage with 100k concurrent tasks or more

  #25412 commented on Jun 16, 2025 • 0 new comments

• [AIR] Utilities to go from Predictor to `BatchPredictor` and `ModelWrapperDeployment`

  #24977 commented on Jun 16, 2025 • 0 new comments

• [Autoscaler][GCP] Autoscaler crashing on GCP with error 404.

  #30050 commented on Jun 16, 2025 • 0 new comments

• Setting some system configs causes Ray to fail to start

  #29841 commented on Jun 16, 2025 • 0 new comments

• [Ray Cluster] Assigning all host GPUs into head node without nvidia.com/gpu present

  #29753 commented on Jun 16, 2025 • 0 new comments

• [gcp] "No such container" error after ray up

  #29671 commented on Jun 16, 2025 • 0 new comments

• [Tune] Passing a handle to grid search cause trials to get stuck in running and pending mode

  #29545 commented on Jun 16, 2025 • 0 new comments

• [Serve] `ServeHandles` fail if GCS crashes before first request

  #29539 commented on Jun 16, 2025 • 0 new comments

• [Core] util.multiprocessing.pool scheduling inefficiencies, blocking behavior in imap and imap_unordered

  #29453 commented on Jun 16, 2025 • 0 new comments

• [Core] inspect_serializability bug - parent object serializable but bound method not

  #29423 commented on Jun 16, 2025 • 0 new comments

• [Core] Ray doesn't shutdown properly on KeyboardInterrupt

  #29384 commented on Jun 16, 2025 • 0 new comments

• [Serve] Unable to upload current working directory

  #29354 commented on Jun 16, 2025 • 0 new comments

• [core][observability] Improving reliability of memory_summary API call

  #29329 commented on Jun 16, 2025 • 0 new comments

• InvalidLocationConstraint Message: The specified location-constraint is not valid for storage option

  #29309 commented on Jun 16, 2025 • 0 new comments

• [Dashboard] A button to shut down the ray cluster from the dashboard UI

  #29208 commented on Jun 16, 2025 • 0 new comments

• [Core] Worker pool didn't prestart num_cpus workers

  #29162 commented on Jun 16, 2025 • 0 new comments

• [core] use proto for oom error / node died error in the frontend

  #28907 commented on Jun 16, 2025 • 0 new comments

• [ray client] surface ray client logs better

  #28890 commented on Jun 16, 2025 • 0 new comments

• [Backlog][Collective] Facilitate NCCL test in ray cluster

  #28860 commented on Jun 16, 2025 • 0 new comments

• [core/k8s/GKE] Ray schedules actors on pods/nodes that are shutting down

  #28852 commented on Jun 16, 2025 • 0 new comments

• [Core] Cannot 'ray list nodes' after setting the environmental variable 'export RAY_ADDRESS="http://127.0.0.1:8265" '

  #28847 commented on Jun 16, 2025 • 0 new comments

• [AIR] [Tune] Don't add random hash to trial id for single trial

  #28830 commented on Jun 16, 2025 • 0 new comments

• [core] Object returned by a generator with num_returns="dynamic" should throw an error if reconstruction fails

  #28688 commented on Jun 16, 2025 • 0 new comments

• [P0] test_submit_cpp_job failed in osx

  #28592 commented on Jun 16, 2025 • 0 new comments

• Unable to override ray's default logging format

  #6965 commented on Jun 16, 2025 • 0 new comments

• MADDPG used onto a MultiEnv does not show learning.

  #6949 commented on Jun 16, 2025 • 0 new comments

• How to throttle process to avoid "UnreconstructableError"

  #6892 commented on Jun 16, 2025 • 0 new comments

• pip install from source requires --editable/-e flag

  #6845 commented on Jun 16, 2025 • 0 new comments

• [scheduling] Default actor lifetime resources (0 CPUs) cause cluster not to be saturated

  #6814 commented on Jun 16, 2025 • 0 new comments

• How to Reduce Memory Usage for Creating Actor?

  #6778 commented on Jun 16, 2025 • 0 new comments

• Reconstruction semantics around failing actor constructor.

  #6768 commented on Jun 16, 2025 • 0 new comments

• [Deploy]Ray on Yarn Deployment

  #6753 commented on Jun 16, 2025 • 0 new comments

• failed on virtualnevironment

  #6735 commented on Jun 16, 2025 • 0 new comments

• Managing memory during long loops

  #6717 commented on Jun 16, 2025 • 0 new comments

• Not able to reproduce speed performance improvements using ray on my machine

  #6716 commented on Jun 16, 2025 • 0 new comments

• [tune] Logs don't sync up to workers on restore

  #6702 commented on Jun 16, 2025 • 0 new comments

• The remote_function.options is not documented.

  #6699 commented on Jun 16, 2025 • 0 new comments

• [tune] More robust checkpoint garbage collection

  #6697 commented on Jun 16, 2025 • 0 new comments

• Fault tolerance to dead actors

  #6670 commented on Jun 16, 2025 • 0 new comments

• ray.wait's num_returns should not fail if num_returns > len(results)

  #6667 commented on Jun 16, 2025 • 0 new comments

• Parallel execution of multiple dataframes by dividing them into sub-frames

  #6640 commented on Jun 16, 2025 • 0 new comments

• Batch Norm example failing under APEX

  #6638 commented on Jun 16, 2025 • 0 new comments

• limiting tensorflow memory failed in actor or function

  #6633 commented on Jun 16, 2025 • 0 new comments

• Remote function is executed in python `exec` with empty local/global will fails

  #6620 commented on Jun 16, 2025 • 0 new comments

• [tune] Estimate timing

  #6618 commented on Jun 16, 2025 • 0 new comments

• [streaming] Add micro batching feature

  #6607 commented on Jun 16, 2025 • 0 new comments

• Proper way of calling a class method in another method

  #7450 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] Provide ability to provide elastic ip when launching cluster

  #7446 commented on Jun 16, 2025 • 0 new comments

• [core] Get IP Address of Actor

  #7431 commented on Jun 16, 2025 • 0 new comments

• What are system requirements for building on Mac OSX

  #7430 commented on Jun 16, 2025 • 0 new comments

• Ray dashboard integration

  #7383 commented on Jun 16, 2025 • 0 new comments

• Do not suggest calling __ray_terminate__ directly

  #7382 commented on Jun 16, 2025 • 0 new comments

• ray.services.get_node_ip_address doesn't work well if there is a local proxy

  #7316 commented on Jun 16, 2025 • 0 new comments

• Provide abstraction/interface to implement resource isolation for custom resources

  #7204 commented on Jun 16, 2025 • 0 new comments

• [cross-language]Problem about cross language data layout

  #7191 commented on Jun 16, 2025 • 0 new comments

• Documentation for connecting to ray cluster could be improved

  #7186 commented on Jun 16, 2025 • 0 new comments

• ray.experimental.queue is very slow

  #7172 commented on Jun 16, 2025 • 0 new comments

• Using asserts for argument checks is probably a bad idea

  #7171 commented on Jun 16, 2025 • 0 new comments

• Ray Issue: The class state is never hold by passing it to the remote function/actors if the class is defined in separate files

  #7160 commented on Jun 16, 2025 • 0 new comments

• Docs on Cython extensions and install requirements

  #7094 commented on Jun 16, 2025 • 0 new comments

• [core] Gets timeout on randomly generated ObjectIDs

  #7074 commented on Jun 16, 2025 • 0 new comments

• Allow remote functions to require running on a fresh worker

  #7059 commented on Jun 16, 2025 • 0 new comments

• How to use Ray with closures?

  #7055 commented on Jun 16, 2025 • 0 new comments

• The project `setup.py` script doesn't install tools needed by `ci/travis/format.sh`

  #6999 commented on Jun 16, 2025 • 0 new comments

• Don't run Java or sanitizer tests when only Python changes.

  #6992 commented on Jun 16, 2025 • 0 new comments

• ray plasma object store connection refused after 24hrs

  #6988 commented on Jun 16, 2025 • 0 new comments

• Sharing in memory

  #6976 commented on Jun 16, 2025 • 0 new comments

• [ray] ray on slurm not respecting memory limits

  #6968 commented on Jun 16, 2025 • 0 new comments

• Package reference should include task & actor APIs

  #6566 commented on Jun 16, 2025 • 0 new comments

• Python Worker class should have proper constructor and destructor.

  #3961 commented on Jun 16, 2025 • 0 new comments

• Should not ignore "AttributeError"

  #3820 commented on Jun 16, 2025 • 0 new comments

• Backend timing statements should be made type safe.

  #3341 commented on Jun 16, 2025 • 0 new comments

• Make it possible to limit memory usage of processes

  #3055 commented on Jun 16, 2025 • 0 new comments

• Task submission from local scheduler client is blocking

  #2940 commented on Jun 16, 2025 • 0 new comments

• Add test for numpy array alignment.

  #2937 commented on Jun 16, 2025 • 0 new comments

• [bug][serve.llm] AssertionError: failed to get the hash of the compiled graph (VLM, batch, TP=2)

  #53824 commented on Jun 16, 2025 • 0 new comments

• Allow ray.get and ray.wait to take in additional argument types

  #2126 commented on Jun 16, 2025 • 0 new comments

• Remove the import thread from the workers and driver.

  #951 commented on Jun 16, 2025 • 0 new comments

• Remote decorator fails on jitted function.

  #593 commented on Jun 16, 2025 • 0 new comments

• Actors do not work properly with subclasses that call super.

  #449 commented on Jun 16, 2025 • 0 new comments

• TypeError: Descriptors cannot not be created directly.

  #36417 commented on Jun 16, 2025 • 0 new comments

• [Tune|RLlib] PBT reward drop - not checkpointing or restoring properly

  #53831 commented on Jun 16, 2025 • 0 new comments

• [rllib] [bug] Official PPO Atari example fails with IndexError

  #53836 commented on Jun 16, 2025 • 0 new comments

• Methods on actors inherited from built-in classes are not visible

  #278 commented on Jun 16, 2025 • 0 new comments

• [Core][StreamingGenerator] `ray.get` will hang when the node on which the streaming task is running fails.

  #47582 commented on Jun 16, 2025 • 0 new comments

• [Serve] `serve.run` can bind the incorrect Application if Deployments have the same name

  #53295 commented on Jun 16, 2025 • 0 new comments

• [Serve] RayServe Pods Stuck in Unready State Causing API Outages

  #53323 commented on Jun 16, 2025 • 0 new comments

• [Serve] Support generics for DeploymentHandle type hints

  #52654 commented on Jun 16, 2025 • 0 new comments

• [Ray Complied Graph] NCCL Internal Error

  #49827 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Checkpointing fails with CUDA GPU learner using the new API stack

  #53793 commented on Jun 16, 2025 • 0 new comments

• How to enable tool calling in serve llm?

  #53795 commented on Jun 16, 2025 • 0 new comments

• Serialization is 20% slower from 0.7.6 -> 0.7.7

  #6551 commented on Jun 16, 2025 • 0 new comments

• [ray] How to write into numpy arrays in shared memory with Ray?

  #6507 commented on Jun 16, 2025 • 0 new comments

• Support for mxnet.ndarray?

  #6494 commented on Jun 16, 2025 • 0 new comments

• [ray] Handle memory pressure more gracefully

  #6458 commented on Jun 16, 2025 • 0 new comments

• Reloading module changes in workers

  #6449 commented on Jun 16, 2025 • 0 new comments

• [tune] [serve] Don't use daemon threads

  #6421 commented on Jun 16, 2025 • 0 new comments

• Terminal freezes after setting @ray.remote(num_gpu=2)

  #6418 commented on Jun 16, 2025 • 0 new comments

• Ray does not preserve requires_grad attribute

  #6405 commented on Jun 16, 2025 • 0 new comments

• Ray over mpi for supercomputers

  #6344 commented on Jun 16, 2025 • 0 new comments

• Support of Ray Decorator for Built in Functions

  #6308 commented on Jun 16, 2025 • 0 new comments

• [docs] Issue on `tune-schedulers.rst`

  #6063 commented on Jun 16, 2025 • 0 new comments

• Can I set priority for my tasks

  #6057 commented on Jun 16, 2025 • 0 new comments

• Avoid putting the redis password in plain text in processlist

  #5872 commented on Jun 16, 2025 • 0 new comments

• Handling `use_pickle=True` with pickle5 serializer and performance regression

  #5856 commented on Jun 16, 2025 • 0 new comments

• Install ray with conda but not pip

  #5511 commented on Jun 16, 2025 • 0 new comments

• [tune] saving mechanism and PBT

  #5312 commented on Jun 16, 2025 • 0 new comments

• Feature request: An API to wait until there are are X resources available

  #5243 commented on Jun 16, 2025 • 0 new comments

• [Feature request] Also expose python function after decorating with ray.remote

  #4981 commented on Jun 16, 2025 • 0 new comments

• Creative action space support: contains method, action interpoalation.

  #4837 commented on Jun 16, 2025 • 0 new comments

• __module__ can be None

  #4758 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] Autoscaler UX Issues

  #4656 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] Add tests that mock endpoints for AWS, GCE

  #4303 commented on Jun 16, 2025 • 0 new comments

• [metrics] Support filtering logs streamed to driver by actor/task

  #12305 commented on Jun 16, 2025 • 0 new comments

• [serve] Support more expressive policies for choosing replicas

  #12296 commented on Jun 16, 2025 • 0 new comments

• [Tune] [PBT] Automatic experiment restart for synch=True

  #12122 commented on Jun 16, 2025 • 0 new comments

• [tune] [wandb] Experiment checkpointing fails with `WandbTrainableMixin`

  #11917 commented on Jun 16, 2025 • 0 new comments

• [tune] quniform distribution

  #11879 commented on Jun 16, 2025 • 0 new comments

• [docs] improve tune distributed tuning guide

  #11681 commented on Jun 16, 2025 • 0 new comments

• [tune] doc should indicate print output

  #11679 commented on Jun 16, 2025 • 0 new comments

• [cli] attach `--tmux` should show parallel command output

  #11678 commented on Jun 16, 2025 • 0 new comments

• [tune] Client API improvements

  #11676 commented on Jun 16, 2025 • 0 new comments

• [cloudpickle] Too much override for cloudpickle, breaks scikit-learn usage

  #11547 commented on Jun 16, 2025 • 0 new comments

• [docs] search results don't link to correct tab

  #11288 commented on Jun 16, 2025 • 0 new comments

• [Autoscaler] Prioritize infeasible bundles and placement group rescheduling

  #11259 commented on Jun 16, 2025 • 0 new comments

• Remove the `remove_after_get` flag

  #10977 commented on Jun 16, 2025 • 0 new comments

• [placement groups] Feasibility Check

  #10913 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] Add unit tests for sdk.py

  #10903 commented on Jun 16, 2025 • 0 new comments

• Cannot call remote instance method of a superclass from within a different instance method of the superclass

  #10899 commented on Jun 16, 2025 • 0 new comments

• Add testing to `commands.py`/`NodeUpdaterThread` level

  #10846 commented on Jun 16, 2025 • 0 new comments

• Treat CPUs as abstract resources

  #10818 commented on Jun 16, 2025 • 0 new comments

• Installing ray on powerpc

  #10774 commented on Jun 16, 2025 • 0 new comments

• [core] [docs] use-cases for Ray's async support

  #10688 commented on Jun 16, 2025 • 0 new comments

• Exceptions and ResourceWarnings on ray.init (Jupyter+offline)

  #10279 commented on Jun 16, 2025 • 0 new comments

• Can CPU resource scheduling be scheduled through Cgroup?

  #10037 commented on Jun 16, 2025 • 0 new comments

• Pre-push hooks allow code to be pushed that fails LINT

  #14367 commented on Jun 16, 2025 • 0 new comments

• __del__ magic method can't access class properties

  #14285 commented on Jun 16, 2025 • 0 new comments

• Failed to load actor due to dependencies not being pickled

  #14284 commented on Jun 16, 2025 • 0 new comments

• optimization: Client blocks on releasing references due to detached actor race condition

  #14137 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] request resources doesn't work with multiple jobs

  #13534 commented on Jun 16, 2025 • 0 new comments

• [Metrics] Custom metrics don't work after calling `ray.shutdown()` followed by `ray.init()`

  #13532 commented on Jun 16, 2025 • 0 new comments

• Unify linting of clang-format and *.proto files

  #13465 commented on Jun 16, 2025 • 0 new comments

• Hang or Deadlock when calling ray.get() inside pytorch Dataset when DataLoader with num_workers >0

  #13407 commented on Jun 16, 2025 • 0 new comments

• [core] Unwanted pickling behaviour when starting remote actor with @propery

  #13365 commented on Jun 16, 2025 • 0 new comments

• Explore Protos as the Ray Client pickle transport (instead of namedtuples)

  #13280 commented on Jun 16, 2025 • 0 new comments

• SIGKILL generates core dumps on some systems

  #13221 commented on Jun 16, 2025 • 0 new comments

• Object store thrashing if it runs ray.get in a non-main thread.

  #12906 commented on Jun 16, 2025 • 0 new comments

• Canonicalize the python lint options

  #12801 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] refactor duplicate code for handling request_resources().

  #12699 commented on Jun 16, 2025 • 0 new comments

• [Dashboard]Profile Actor Button Not Working

  #12668 commented on Jun 16, 2025 • 0 new comments

• [core] bytearray is parsed as bytes in remote function

  #12648 commented on Jun 16, 2025 • 0 new comments

• ValueError: After taking into account object store and redis memory usage, the amount of memory on this node available for tasks and actors (-0.01 GB) is less than 0% of total. You can adjust these settings with ray.init(memory=<bytes>, object store memory=<bytes>

  #12561 commented on Jun 16, 2025 • 0 new comments

• Ray grinds to a halt if both PyTorch and TensorFlow are installed

  #12467 commented on Jun 16, 2025 • 0 new comments

• Ray does not handle MIG devices

  #12413 commented on Jun 16, 2025 • 0 new comments

• [tune] progress reporter should limit table to 80char

  #12374 commented on Jun 16, 2025 • 0 new comments

• [serve] Distributed Tracing Support in Serve

  #12320 commented on Jun 16, 2025 • 0 new comments

• [metrics] Replace ray timeline with distributed tracing

  #12315 commented on Jun 16, 2025 • 0 new comments

• Windows debugging on gdb does not work

  #9827 commented on Jun 16, 2025 • 0 new comments

• [tune] unify run() and run_experiments()

  #8127 commented on Jun 16, 2025 • 0 new comments

• [ui] More metadata for the task timeline

  #8050 commented on Jun 16, 2025 • 0 new comments

• [tune] Support for config to (optionally) be an argparse.Namespace?

  #8006 commented on Jun 16, 2025 • 0 new comments

• [tune] Resource Allocation UX

  #7968 commented on Jun 16, 2025 • 0 new comments

• `pandas has no attribute 'compat'` Deserialization bug when running tasks very rarely

  #7879 commented on Jun 16, 2025 • 0 new comments

• "Lost reference to actor" when returning actor handle from actor

  #7815 commented on Jun 16, 2025 • 0 new comments

• Ray has both ray.util and ray.utils, which is confusing.

  #7787 commented on Jun 16, 2025 • 0 new comments

• Provide more scheduling algorithms for actors/tasks

  #7723 commented on Jun 16, 2025 • 0 new comments

• [ray] Object store shared memory numpy leak in worker loop

  #7653 commented on Jun 16, 2025 • 0 new comments

• Ray processes on slave node become defunct when the head node is restarted/stopped

  #7651 commented on Jun 16, 2025 • 0 new comments

• Relax python version match requirement when joining a cluster

  #7648 commented on Jun 16, 2025 • 0 new comments

• Does ray workers could share the same tf.sess?

  #7646 commented on Jun 16, 2025 • 0 new comments

• About model configuration.

  #7644 commented on Jun 16, 2025 • 0 new comments

• Probable race condition

  #7617 commented on Jun 16, 2025 • 0 new comments

• Recursion with pickling in ray.init with py3.5

  #7605 commented on Jun 16, 2025 • 0 new comments

• Is it possible to create process inside ray Actor?

  #7578 commented on Jun 16, 2025 • 0 new comments

• Why seems getting from local object store not faster than getting from remote object store?

  #7575 commented on Jun 16, 2025 • 0 new comments

• [util.multiprocessing] Unable to pass Queue to pool.apply_async

  #7561 commented on Jun 16, 2025 • 0 new comments

• Keyword arguments should be keyword only arguments in the Ray API

  #7548 commented on Jun 16, 2025 • 0 new comments

• [Pool] About using ray.util.multiprocessing import Pool

  #7542 commented on Jun 16, 2025 • 0 new comments

• Reporting Reward Breakdowns

  #7518 commented on Jun 16, 2025 • 0 new comments

• [config] Introduce a configuration library for unified configuration code

  #7485 commented on Jun 16, 2025 • 0 new comments

• [util.multiprocessing] Support generators

  #9712 commented on Jun 16, 2025 • 0 new comments

• [Core] A ray.remote flag for nested object ID gathering in task arguments.

  #9489 commented on Jun 16, 2025 • 0 new comments

• [docs] ray up <config.xml> --help does not show help

  #9455 commented on Jun 16, 2025 • 0 new comments

• [docs] Document how to use conda environments with the autoscaler

  #9199 commented on Jun 16, 2025 • 0 new comments

• [ray] Visualize Ray dashboard locally/offline

  #9095 commented on Jun 16, 2025 • 0 new comments

• Confusing RedisError when many threads are used

  #9083 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] Check failed: _s.ok() Heartbeat failed: NotImplemented

  #8883 commented on Jun 16, 2025 • 0 new comments

• Can't parallelize non-pickable function with initializer in Pool

  #8876 commented on Jun 16, 2025 • 0 new comments

• DQN Minibatch Option

  #8870 commented on Jun 16, 2025 • 0 new comments

• tune: module 'tensorflow' has no attribute __version__ in Ray Trainable since v0.7.7

  #8729 commented on Jun 16, 2025 • 0 new comments

• Blank redis-password gives wrong message to add node

  #8629 commented on Jun 16, 2025 • 0 new comments

• absl.logging inside remote tasks does not get printed

  #8625 commented on Jun 16, 2025 • 0 new comments

• Ability to select a disk for ray workers

  #8607 commented on Jun 16, 2025 • 0 new comments

• Invalid iterator dereference in TestReconstructionChain (fails in debug mode)

  #8587 commented on Jun 16, 2025 • 0 new comments

• Can't pickle CudnnModule objects

  #8569 commented on Jun 16, 2025 • 0 new comments

• Reducing unnecessary process overhead in practice

  #8522 commented on Jun 16, 2025 • 0 new comments

• [tune]Error in BOHB perhaps caused by different trainable instances running in the same Trial ???

  #8455 commented on Jun 16, 2025 • 0 new comments

• incompatible with 'msgpack_numpy.patch()' function

  #8409 commented on Jun 16, 2025 • 0 new comments

• Error connecting to Redis server at 127.0.0.1:35709

  #8389 commented on Jun 16, 2025 • 0 new comments

• Error while shutting down Ray

  #8385 commented on Jun 16, 2025 • 0 new comments

• [ray] Pyarmor compatibility

  #8365 commented on Jun 16, 2025 • 0 new comments

• [ray] Can RAY pause and continue tasks distributed to the cluster's nodes?

  #8263 commented on Jun 16, 2025 • 0 new comments

• [Bug] Potential deadlock in task scheduling algorithm for placement group resources.

  #20051 commented on Jun 16, 2025 • 0 new comments

• [Object Spilling] Plasma store probably doesn't respect the max shm size.

  #14145 commented on Jun 16, 2025 • 0 new comments

• Latent bugs in command_runner.py

  #14139 commented on Jun 16, 2025 • 0 new comments

• [rllib] undocumented behavior of timers/* in progress.csv

  #14052 commented on Jun 16, 2025 • 0 new comments

• Graceful Placement Group Removal

  #14045 commented on Jun 16, 2025 • 0 new comments

• Improve Docker manual setup document

  #14030 commented on Jun 16, 2025 • 0 new comments

• [UX] Allow passing CPU and GPU to actor and task resources.

  #13996 commented on Jun 16, 2025 • 0 new comments

• Remove cluster_synced_files and file_mounts_sync_continuously

  #13967 commented on Jun 16, 2025 • 0 new comments

• [Object Spilling] Allow to specify max_disk_usage for file system spilling.

  #13960 commented on Jun 16, 2025 • 0 new comments

• [Dashboard] add actor detail to experimental dashboard

  #13875 commented on Jun 16, 2025 • 0 new comments

• ray.put() slows down over time.

  #13612 commented on Jun 16, 2025 • 0 new comments

• [rllib]Action masking with tuple action space

  #13592 commented on Jun 16, 2025 • 0 new comments

• [dask-on-ray] Remove internal Dask API dependencies from the Dask-on-Ray scheduler.

  #13560 commented on Jun 16, 2025 • 0 new comments

• [core] GCS doesn't always cancel worker leases for killed actors

  #13545 commented on Jun 16, 2025 • 0 new comments

• test_autoscaling_policy.py prints out huge pile of JsonErrors

  #13433 commented on Jun 16, 2025 • 0 new comments

• Remove the RAY_CLIENT_MODE flag now that we don't need it

  #13279 commented on Jun 16, 2025 • 0 new comments

• [Core] Make CoreWorker more unit-testable

  #13268 commented on Jun 16, 2025 • 0 new comments

• Test S3 object spilling on multiple nodes with big data (streaming shuffle)

  #13222 commented on Jun 16, 2025 • 0 new comments

• [core] RAY_HOME path is hardcoded

  #13168 commented on Jun 16, 2025 • 0 new comments

• [Plasma Store]PlasmaClient::Get() return Status::OK() when timeout

  #12995 commented on Jun 16, 2025 • 0 new comments

• Add dashboard to bazel target to avoid running manual build commands

  #12956 commented on Jun 16, 2025 • 0 new comments

• Improve dashboard not found exception

  #12955 commented on Jun 16, 2025 • 0 new comments

• Cannot save training episodes: "TypeError: Object of type ndarray is not JSON serializable"

  #12951 commented on Jun 16, 2025 • 0 new comments

• num_cpus not handled correctly when function has a Queue argument

  #14863 commented on Jun 16, 2025 • 0 new comments

• Make rolling update batch size configurable

  #14853 commented on Jun 16, 2025 • 0 new comments

• Typed handle to deployments

  #14810 commented on Jun 16, 2025 • 0 new comments

• [Core] Docs - run data processing examples in CI

  #14769 commented on Jun 16, 2025 • 0 new comments

• [core] The remote function has been exported 100 times..

  #14730 commented on Jun 16, 2025 • 0 new comments

• Support `ray status CLUSTER.YAML`

  #14549 commented on Jun 16, 2025 • 0 new comments

• Support decoupling task/actor interfaces from implementation

  #14529 commented on Jun 16, 2025 • 0 new comments

• Support specifying container images in runtime_env

  #14528 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] Error message not being cleared when autoscaler recovers

  #14494 commented on Jun 16, 2025 • 0 new comments

• [Docs] [tune] WanDB + Ray Integration a bit unclear from the docs

  #14478 commented on Jun 16, 2025 • 0 new comments

• [tune] TBXLoggerCallback not creating necessary directory

  #14437 commented on Jun 16, 2025 • 0 new comments

• [autoscaler][interface] Per-node-type docker configs

  #14418 commented on Jun 16, 2025 • 0 new comments

• [metrics] Add metrics for debugging Dask-on-Ray

  #14372 commented on Jun 16, 2025 • 0 new comments

• [metrics] Report metrics to be used for debugging load balancing issues

  #14369 commented on Jun 16, 2025 • 0 new comments

• [metrics] Remove unused or unnecessary metrics.

  #14366 commented on Jun 16, 2025 • 0 new comments

• When the node is crashed, logs are not accessible.

  #14307 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] SSH command errors aren't written to monitor.out

  #14298 commented on Jun 16, 2025 • 0 new comments

• [dashboard] Add resource usage/availability to the dashboard

  #14292 commented on Jun 16, 2025 • 0 new comments

• [Core] Fix ray::Status <--> gRPC status interplay.

  #14278 commented on Jun 16, 2025 • 0 new comments

• updating worker nodes show as healthy

  #14232 commented on Jun 16, 2025 • 0 new comments

• [Object Spilling] Use subdirectories to avoid large top level inodes for file spilling

  #14166 commented on Jun 16, 2025 • 0 new comments

• [tune] Stack Traces with Function API are really hard to parse

  #14162 commented on Jun 16, 2025 • 0 new comments

• [Object Spilling] Improve Read throughput

  #12950 commented on Jun 16, 2025 • 0 new comments

• [tune] Restarted Trials Use Incorrect Command When Multiple Commands Run on Cluster/Runtime

  #12048 commented on Jun 16, 2025 • 0 new comments

• [Object spilling] Move LocalObjectManager into the plasma store

  #12042 commented on Jun 16, 2025 • 0 new comments

• [Object spilling] Improve OutOfMemory handling through better memory bookkeeping in plasma store

  #12040 commented on Jun 16, 2025 • 0 new comments

• [Object Spilling] Use compression to reduce IO cost.

  #11992 commented on Jun 16, 2025 • 0 new comments

• [Tune] Add more custom Error Types

  #11871 commented on Jun 16, 2025 • 0 new comments

• Tune report histograms

  #11797 commented on Jun 16, 2025 • 0 new comments

• [Feature Request] [Tune] Add a special 'evaluation-step' flag to avoid unnecessary lengthy evaluations

  #11725 commented on Jun 16, 2025 • 0 new comments

• Unable to create ActorHandle for already created inherited classes object list [java][ray]

  #11715 commented on Jun 16, 2025 • 0 new comments

• Socket connections from GCS stuck in TIME_WAIT after actor death

  #11713 commented on Jun 16, 2025 • 0 new comments

• [docs] tutorial for autoscaling (really basic version)

  #11680 commented on Jun 16, 2025 • 0 new comments

• [flaky] test_multi_node/2 is flaky

  #11663 commented on Jun 16, 2025 • 0 new comments

• [flaky] test_object_manager is flaky

  #11661 commented on Jun 16, 2025 • 0 new comments

• [Core] Reduce the Redis connection per worker.

  #11655 commented on Jun 16, 2025 • 0 new comments

• [flaky] gcs_server test is flaky

  #11640 commented on Jun 16, 2025 • 0 new comments

• Use Pathlib instead of strings in Autoscaler

  #11633 commented on Jun 16, 2025 • 0 new comments

• [tune] PopulationBasedTraining and Tensorboard HPARAMS

  #11612 commented on Jun 16, 2025 • 0 new comments

• AWS Security group rule issue

  #11601 commented on Jun 16, 2025 • 0 new comments

• [dask] Parquet write fails if directory does not exist in advance

  #11566 commented on Jun 16, 2025 • 0 new comments

• [dask] Object store fills up too quickly in simple processing script

  #11565 commented on Jun 16, 2025 • 0 new comments

• [dask/tune] Provide an example of using Dask on Ray with Tune

  #11564 commented on Jun 16, 2025 • 0 new comments

• [tune] tutorial should indicate specific library version that we've tested against.

  #11540 commented on Jun 16, 2025 • 0 new comments

• [Core] Raylet can schedule tasks from a dead driver.

  #11520 commented on Jun 16, 2025 • 0 new comments

• Startup log use autoscaler_log.out / err instead of monitor.log

  #12884 commented on Jun 16, 2025 • 0 new comments

• [New scheduler] Don't assume 1-CPU tasks are feasible

  #12870 commented on Jun 16, 2025 • 0 new comments

• Turn on Test_reference_counting

  #12849 commented on Jun 16, 2025 • 0 new comments

• [Core] Locality-aware leasing: Milestone 3 - Spillback

  #12815 commented on Jun 16, 2025 • 0 new comments

• [Autoscaler] Refactor bin packing routines in autoscaler for code clarity

  #12723 commented on Jun 16, 2025 • 0 new comments

• [Core] Ray.get(timeout=0) doesn't work

  #12680 commented on Jun 16, 2025 • 0 new comments

• [core] Is starvation possible for multi-driver on the same cluster?

  #12667 commented on Jun 16, 2025 • 0 new comments

• GCS server ip error

  #12639 commented on Jun 16, 2025 • 0 new comments

• [core] Support detached/GCS owned objects

  #12635 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] respect max_workers per node type when terminating nodes

  #12634 commented on Jun 16, 2025 • 0 new comments

• [Cluster launcher] Command runner logs are improperly quoted when logged

  #12631 commented on Jun 16, 2025 • 0 new comments

• permissions on rsync'd files are incorrect on worker nodes, results in inability to update workers

  #12630 commented on Jun 16, 2025 • 0 new comments

• [tune] Full experiment checkpointing doesn't work with PBT

  #12558 commented on Jun 16, 2025 • 0 new comments

• New workers are started slowly on a node if running workers >= `num_cpus`

  #12525 commented on Jun 16, 2025 • 0 new comments

• [tune] get_checkpoint_paths fails due to glob command for .tune_metadata file

  #12453 commented on Jun 16, 2025 • 0 new comments

• [New scheduler] Implement dynamic resources

  #12433 commented on Jun 16, 2025 • 0 new comments

• [metrics] Investigate tracing visualization tools

  #12314 commented on Jun 16, 2025 • 0 new comments

• [metrics] Utility to easily configure logging for a Ray job/actor/task

  #12306 commented on Jun 16, 2025 • 0 new comments

• `ray dashboard` throws bad exception

  #12246 commented on Jun 16, 2025 • 0 new comments

• [Object Spilling] Tune S3 performance + Add unit tests with moto3

  #12232 commented on Jun 16, 2025 • 0 new comments

• Duplicated IDs are generated

  #12197 commented on Jun 16, 2025 • 0 new comments

• [tune/logging] Warning for Tune

  #12140 commented on Jun 16, 2025 • 0 new comments

• [C++ API] Support cross-lang API with Python/Java

  #18149 commented on Jun 16, 2025 • 0 new comments

• [helm][kubernetes][test] Add formatting tests for Helm chart

  #18125 commented on Jun 16, 2025 • 0 new comments

• [workflows] Better message when not init'ed

  #18121 commented on Jun 16, 2025 • 0 new comments

• resource config is not respected in head_start_ray_commands in cluster.yaml

  #18097 commented on Jun 16, 2025 • 0 new comments

• [Dask-on-Ray] Propagate Dask-on-Ray scheduler config to (rest of) cluster

  #17943 commented on Jun 16, 2025 • 0 new comments

• [core] PlacementGroup should be no op for local_mode=True

  #17937 commented on Jun 16, 2025 • 0 new comments

• Enhance document on Java API

  #17820 commented on Jun 16, 2025 • 0 new comments

• [Object Spilling] Remove the spilled directory upon Sigterm for ray start

  #17790 commented on Jun 16, 2025 • 0 new comments

• [C++ API] Support non-global named actor

  #17734 commented on Jun 16, 2025 • 0 new comments

• Cleanup stats/metrics.h

  #17679 commented on Jun 16, 2025 • 0 new comments

• workflow cli to manage all jobs

  #17672 commented on Jun 16, 2025 • 0 new comments

• [docs] Tutorial on Pytorch Lightning needs rearranging

  #17611 commented on Jun 16, 2025 • 0 new comments

• [Serve] Helper functions that are written below the actor class don't work

  #17590 commented on Jun 16, 2025 • 0 new comments

• Fix circular dependence in workflow's code

  #17445 commented on Jun 16, 2025 • 0 new comments

• [lineage] Support lineage reconstruction for borrowed ObjectRefs

  #17380 commented on Jun 16, 2025 • 0 new comments

• Errors during scaling cluster

  #17292 commented on Jun 16, 2025 • 0 new comments

• Trial is being repeated with the exact same results

  #17257 commented on Jun 16, 2025 • 0 new comments

• [RFC][Placement groups] Allow tasks to acquire resources in addition to placement group bundle

  #17229 commented on Jun 16, 2025 • 0 new comments

• runtime env in workflow

  #16992 commented on Jun 16, 2025 • 0 new comments

• [autoscaler][core] Safe node termination

  #16975 commented on Jun 16, 2025 • 0 new comments

• [Ray Client] [Usability] Help users spot bandwidth bounded workload

  #16966 commented on Jun 16, 2025 • 0 new comments

• [docker][Clusters][autoscaler][local] Can't connect to cluster when using docker with ray cluster launcher

  #16961 commented on Jun 16, 2025 • 0 new comments

• [Feature] rllib + tune metric logging selection

  #19816 commented on Jun 16, 2025 • 0 new comments

• [RLlib] [documentation] clarify postprocess_fn usage in our doc

  #19648 commented on Jun 16, 2025 • 0 new comments

• [Feature] [runtime env] Clean up the command arguments in raylet args

  #19448 commented on Jun 16, 2025 • 0 new comments

• [client] better error message when failing to connect with client

  #19371 commented on Jun 16, 2025 • 0 new comments

• [SGD] Document best practices for Pipeline epochs

  #19323 commented on Jun 16, 2025 • 0 new comments

• [workflow] scan_prefix with pages/as geneartor

  #19234 commented on Jun 16, 2025 • 0 new comments

• [Core][usability] Improve Ray cluster start up time

  #19215 commented on Jun 16, 2025 • 0 new comments

• [Serve] Don't use `ray.wait()` to drain tracking refs in handle

  #19158 commented on Jun 16, 2025 • 0 new comments

• Unify internal configs & common datastructures

  #19152 commented on Jun 16, 2025 • 0 new comments

• Clean up EndpointState

  #19148 commented on Jun 16, 2025 • 0 new comments

• [Core][Feature] use clang-tidy/format to block usage of std::getenv

  #18894 commented on Jun 16, 2025 • 0 new comments

• [Feature][workflow] Namespace for workflow

  #18818 commented on Jun 16, 2025 • 0 new comments

• [Feature][workflow] Resource limit for workflow job

  #18780 commented on Jun 16, 2025 • 0 new comments

• [Bug] Exception in task leads to truncated error message

  #18699 commented on Jun 16, 2025 • 0 new comments

• [Bug] Logging config is not propagated to driver

  #18660 commented on Jun 16, 2025 • 0 new comments

• Enable copy/paste to get correct command for connecting to Ray client

  #18513 commented on Jun 16, 2025 • 0 new comments

• Ray client suppresses error messages

  #18512 commented on Jun 16, 2025 • 0 new comments

• [serve] Feature request: timeout or max_retries to limit the time spent waiting for a deployment to complete

  #18432 commented on Jun 16, 2025 • 0 new comments

• Add workflow.current_step_uuid() function

  #18356 commented on Jun 16, 2025 • 0 new comments

• [Shuffle] non streaming shuffle 5000 partitions seem to reach the scalability limit

  #18333 commented on Jun 16, 2025 • 0 new comments

• [tune] atari-impala-large.yaml does not finish gracefully

  #18325 commented on Jun 16, 2025 • 0 new comments

• [runtime env] eagerly install for task/actor level

  #18160 commented on Jun 16, 2025 • 0 new comments

• [cli] Support redis password for all ray commands

  #16921 commented on Jun 16, 2025 • 0 new comments

• [client][core] Have Unified `register_serializer` interface

  #15486 commented on Jun 16, 2025 • 0 new comments

• [Job submission] Monitor driver

  #15480 commented on Jun 16, 2025 • 0 new comments

• [Job submission] Java support

  #15479 commented on Jun 16, 2025 • 0 new comments

• [Job submission] Basic drop job feature

  #15478 commented on Jun 16, 2025 • 0 new comments

• Ray memory size and object store size not correct on k8s

  #15463 commented on Jun 16, 2025 • 0 new comments

• Ray status not report correctly after node crashed

  #15459 commented on Jun 16, 2025 • 0 new comments

• Async actor method hang

  #15437 commented on Jun 16, 2025 • 0 new comments

• [client] python packages version mismatch fail silently

  #15407 commented on Jun 16, 2025 • 0 new comments

• [cluster] Make node_ip_address work throughout

  #15239 commented on Jun 16, 2025 • 0 new comments

• [autoscaler][docs] Explain how the `ray_bootstrap_config` is generated

  #15232 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] Don't autofill `setup_commands` if head/worker `setup_commands` are used

  #15231 commented on Jun 16, 2025 • 0 new comments

• [Core] Add gRPC streaming support.

  #15219 commented on Jun 16, 2025 • 0 new comments

• Optimise for num_workers stucks in the infinite loop

  #15168 commented on Jun 16, 2025 • 0 new comments

• Ray dies without a proper error message - "Killed", might have to do with pandas

  #15165 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] Simplify Custom ObjectStore Size

  #15147 commented on Jun 16, 2025 • 0 new comments

• [Core] Periodical runner can cause heap-use-after-free

  #15141 commented on Jun 16, 2025 • 0 new comments

• Metric tag keys type inference (Tuple To String)

  #15130 commented on Jun 16, 2025 • 0 new comments

• Actor task hangs after actor crashes with max_task_retries=0

  #15045 commented on Jun 16, 2025 • 0 new comments

• AlphaZero torch model doesn't support cuda, only cpu

  #14970 commented on Jun 16, 2025 • 0 new comments

• [Autoscaler] AWS setup commands hardcodes pip

  #14963 commented on Jun 16, 2025 • 0 new comments

• Support "dry runs" for deploy() operations

  #14936 commented on Jun 16, 2025 • 0 new comments

• [Object Spilling] Failing objects that fail to restore many times.

  #14921 commented on Jun 16, 2025 • 0 new comments

• [Core] [runtime env] Use portable hash function for runtime_env_hash

  #16821 commented on Jun 16, 2025 • 0 new comments

• [runtime env] Support rescheduling tasks when runtime env creation failed.

  #16800 commented on Jun 16, 2025 • 0 new comments

• Priority scheduling of jobs

  #16782 commented on Jun 16, 2025 • 0 new comments

• [C++ API] Completed object reference counting support

  #16702 commented on Jun 16, 2025 • 0 new comments

• [Core] Programmatic way to access pending tasks for an actor?

  #16641 commented on Jun 16, 2025 • 0 new comments

• [Core] Erroneous check for size_t underflow

  #16626 commented on Jun 16, 2025 • 0 new comments

• [Core] Standardize Timestamps across codebase

  #16510 commented on Jun 16, 2025 • 0 new comments

• [test][MLDataset] Fix test_from_modin

  #16357 commented on Jun 16, 2025 • 0 new comments

• Example for tuning layer count, dropout probabilities with Transformers

  #16340 commented on Jun 16, 2025 • 0 new comments

• Ray started in local mode doesn't restore environment variables after shutdown

  #16132 commented on Jun 16, 2025 • 0 new comments

• [core] ray.remote hides the docstring of the decorated class

  #15877 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] support rsync option `--include`

  #15859 commented on Jun 16, 2025 • 0 new comments

• [Placement Group] The bundle_reservation_check_func breaks load code from local

  #15840 commented on Jun 16, 2025 • 0 new comments

• Contributor docs don't mention running tests via bazel

  #15833 commented on Jun 16, 2025 • 0 new comments

• [docs] should actor methods always have num_returns value?

  #15818 commented on Jun 16, 2025 • 0 new comments

• [rfc] Support `ray[aws,gcp,azure]` as an install target

  #15725 commented on Jun 16, 2025 • 0 new comments

• [rllib] Error while using "count_steps_by": "agent_steps" and misleading documentation

  #15708 commented on Jun 16, 2025 • 0 new comments

• Ray duplicate data from GPU to CPU when placing an actor on GPU

  #15692 commented on Jun 16, 2025 • 0 new comments

• [kubernetes] ModuleNotFoundError when executing a task on a remote cluster

  #15668 commented on Jun 16, 2025 • 0 new comments

• [cross_language] Support Python dictionaries

  #15569 commented on Jun 16, 2025 • 0 new comments

• [core] detached actor logs are not streamed to successive clients

  #15549 commented on Jun 16, 2025 • 0 new comments

• Serve Deployment with Reload Option

  #15505 commented on Jun 16, 2025 • 0 new comments

• Building an executable using Ray and Cx_freeze

  #42101 commented on Jun 16, 2025 • 0 new comments

• RichProgressBar in PyTorch Lightning only show progress at the very end

  #42091 commented on Jun 16, 2025 • 0 new comments

• [<Ray component: Core|RLlib|etc...>] Channel errore

  #42089 commented on Jun 16, 2025 • 0 new comments

• [RLlib] When setting `config.environment(normalize_actions=False)` and using CQL, it raises an error: `AttributeError: 'TorchDiagGaussian' object has no attribute 'sample_logp'`.

  #42064 commented on Jun 16, 2025 • 0 new comments

• [Workflow] get_metadata() returns RUNNING instead of RESUMABLE status

  #41980 commented on Jun 16, 2025 • 0 new comments

• Ray IDs vs endianness?

  #41961 commented on Jun 16, 2025 • 0 new comments

• "RaySystemError: System error: Unknown error"

  #41786 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Value error while running DQN

  #41559 commented on Jun 16, 2025 • 0 new comments

• Is there an error in the interface name " cdef cppclass CFunctionDescriptorInterface "ray::CFunctionDescriptorInterface "?

  #41553 commented on Jun 16, 2025 • 0 new comments

• Saving XGBoost model with json extension

  #41374 commented on Jun 16, 2025 • 0 new comments

• [data] date32 and datetime64 handling should be the same

  #41358 commented on Jun 16, 2025 • 0 new comments

• [RLlib] User guides are not ordered

  #41340 commented on Jun 16, 2025 • 0 new comments

• error installing library

  #41223 commented on Jun 16, 2025 • 0 new comments

• [core][state][dashboard] Better tasks info GC control at GCS

  #41142 commented on Jun 16, 2025 • 0 new comments

• [RLLib] External simulator: mean episode reward is NaN due to done not set

  #40954 commented on Jun 16, 2025 • 0 new comments

• [Core] - Cannot install in tiny core linux

  #40832 commented on Jun 16, 2025 • 0 new comments

• [Tune|RLlib] Add error-tolerant version of PB2

  #40787 commented on Jun 16, 2025 • 0 new comments

• NCCL Proxy Call to rank 1 failed - on Cloud VM Docker setup for huggingface distributed ray train script

  #40758 commented on Jun 16, 2025 • 0 new comments

• [Ray Train] - Add Options to Save Last checkpoint in Ray Train Checkpointing Config

  #40503 commented on Jun 16, 2025 • 0 new comments

• ray.init() can sometimes hang with a limited range specified for --worker-port-list

  #40497 commented on Jun 16, 2025 • 0 new comments

• [Core] Dead session not closed

  #40482 commented on Jun 16, 2025 • 0 new comments

• [RLlib][MBMPO] The algorithm does not learn as intended.

  #40400 commented on Jun 16, 2025 • 0 new comments

• [Data] Add `delete_dir_contents` parameter to `FileDatasink`

  #44794 commented on Jun 16, 2025 • 0 new comments

• [RLlib] PPO and framework=tf / issue with latest tensorflow 2.16.1

  #44675 commented on Jun 16, 2025 • 0 new comments

• [RLlib] PPO reset_config() AttributeError: 'dict' object has no attribute '_enable_new_api_stack'

  #44506 commented on Jun 16, 2025 • 0 new comments

• [RLlib] ReplayBuffer doesnt work with zero_init_states False when store rnn sequence

  #44383 commented on Jun 16, 2025 • 0 new comments

• [Cluster, YARN with Skein] Ray cluster keeps crashing when running on YARN via Skein

  #44112 commented on Jun 16, 2025 • 0 new comments

• [Ray Core] Ray nightly GPU docker image broken on NVIDIA V100 GPUs on AWS

  #43565 commented on Jun 16, 2025 • 0 new comments

• Using RNN for RL

  #43420 commented on Jun 16, 2025 • 0 new comments

• from ray.rllib.agents.registry import get_trainer_class ModuleNotFoundError: No module named 'ray.rllib.agents'

  #43310 commented on Jun 16, 2025 • 0 new comments

• WARNING deprecation.py:50 -- DeprecationWarning: `ray.rllib.execution.train_ops.multi_gpu_train_one_step` has been deprecated. This will raise an error in the future!

  #43250 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Build_for_inference() in env_runner_v2.py created empty state_out_1 and lead to failure of initiation

  #42978 commented on Jun 16, 2025 • 0 new comments

• Core: ray.remote raises ValueError when used on torch IterableDataset

  #42914 commented on Jun 16, 2025 • 0 new comments

• Core: Join zombie subprocesses after task completion

  #42913 commented on Jun 16, 2025 • 0 new comments

• [Core] SIGSEGV when running Ray

  #42868 commented on Jun 16, 2025 • 0 new comments

• [Core] Serialisation does not work with classes with `__init_subclass__`

  #42823 commented on Jun 16, 2025 • 0 new comments

• Problem with YOLOv8 Hyperparameters tuning

  #42770 commented on Jun 16, 2025 • 0 new comments

• [RLLIB] Passing configuration to Custom Environment in rllib is giving an error

  #42753 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Algorithms ES, A3C are deprecated and replacement does not exist in python package

  #42579 commented on Jun 16, 2025 • 0 new comments

• [<Ray component: Core|RLlib|etc...>] Inite state of attention_net.py is empty

  #42569 commented on Jun 16, 2025 • 0 new comments

• [<Ray component: Core|RLlib|etc...>] KeyError with RNN

  #42501 commented on Jun 16, 2025 • 0 new comments

• [RLlib] gpu cannot enable

  #42388 commented on Jun 16, 2025 • 0 new comments

• [<Ray component: Core|RLlib|etc...>] reslink in model

  #42333 commented on Jun 16, 2025 • 0 new comments

• [RLlib] shape [] in Box action space not supported.

  #42199 commented on Jun 16, 2025 • 0 new comments

• [Tune] Support for new algorithm: Cost-Aware Pareto Region Bayesian Search (CARBS).

  #40356 commented on Jun 16, 2025 • 0 new comments

• latest ray microbenchmark fails

  #38758 commented on Jun 16, 2025 • 0 new comments

• Ray Memory Usage Keeps Increasing even after Manual Garbage Collection

  #38730 commented on Jun 16, 2025 • 0 new comments

• [docs] Document Tune/Train placement group

  #38706 commented on Jun 16, 2025 • 0 new comments

• ray/RLlib/offline/estimators

  #38357 commented on Jun 16, 2025 • 0 new comments

• [Core] gcs_server Failed accept4: Too many open files

  #38248 commented on Jun 16, 2025 • 0 new comments

• [Core] Segfault from fibers when using streaming/dynamic generator (only happening from test_streaming_generator_exception)

  #38167 commented on Jun 16, 2025 • 0 new comments

• [RLlib] RL module and PPO implementation

  #38012 commented on Jun 16, 2025 • 0 new comments

• [RLlib] ray 2.6 relies on tf.bool which does not exist in tensorflow 2.13

  #37895 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Sampler takes first step before next batch is requested

  #37893 commented on Jun 16, 2025 • 0 new comments

• [Data] Ray 2.6 created a breaking change in the index of a Modin DataFrame

  #37771 commented on Jun 16, 2025 • 0 new comments

• [Ray-Java client] Call actor report 'No module named' with py script

  #37600 commented on Jun 16, 2025 • 0 new comments

• [Core] Ray cpp example, if not call ray::Shutdown when exit, will cause segment fault.

  #37596 commented on Jun 16, 2025 • 0 new comments

• RLLib: Training Rllib-DDPG with custom environment leads error in Inference.

  #37242 commented on Jun 16, 2025 • 0 new comments

• [<Ray component: autoscaler>] _load_kubernetes_defaults_config function is not yet made

  #37033 commented on Jun 16, 2025 • 0 new comments

• [Core] Activate Ray tracing casue error when calling actor method with decorator with wraps. (i propose the possible solution)

  #36891 commented on Jun 16, 2025 • 0 new comments

• [Core] No dependency on setuptools results in broken build

  #36742 commented on Jun 16, 2025 • 0 new comments

• [data] Report actual task time and object sizes in Dataset.stats()

  #36671 commented on Jun 16, 2025 • 0 new comments

• [CI][Docs] Example in Train FAQ is flakey

  #36399 commented on Jun 16, 2025 • 0 new comments

• Be consistent on whether or not you include a dot at the end of a bullet list element.

  #36308 commented on Jun 16, 2025 • 0 new comments

• [Core] ray.put and ray.get extremely slow with polars frames

  #36068 commented on Jun 16, 2025 • 0 new comments

• [Ray Core] There is a Exception error message bug which convert byte array to String.

  #35880 commented on Jun 16, 2025 • 0 new comments

• System error: Ray has not been started yet. You can start Ray with 'ray.init()'

  #35592 commented on Jun 16, 2025 • 0 new comments

• [Workflow] Incorrectly set max_calls in options

  #40252 commented on Jun 16, 2025 • 0 new comments

• [PPOConfig] Utilising new API/models without matching documentation

  #40201 commented on Jun 16, 2025 • 0 new comments

• [Rllib] Tune locks up when attempting to create an rllib algorithm in a trainable

  #40015 commented on Jun 16, 2025 • 0 new comments

• [Tune/Air] Memory Leak when using WandbLoggerCallback with Population Based Tuning

  #40014 commented on Jun 16, 2025 • 0 new comments

• [RLlib] TD3/DDPG doesn't seem to respect action space bounds (at least initially)?

  #40002 commented on Jun 16, 2025 • 0 new comments

• [RLLIB] Issue with AlphaZero algorithm Stateless CartPole

  #39937 commented on Jun 16, 2025 • 0 new comments

• [RLLIB] Error in executing StatelessCartPole environment with AlphaZero

  #39862 commented on Jun 16, 2025 • 0 new comments

• Allow train_loop_config to be a dataclass / pydantic model

  #39824 commented on Jun 16, 2025 • 0 new comments

• [Core] ResolutionImpossible - Test requirements appear to not fit versions

  #39782 commented on Jun 16, 2025 • 0 new comments

• Job history is lost when Ray cluster is restarted (via kuberay)

  #39764 commented on Jun 16, 2025 • 0 new comments

• Ray::Tune::Logger::Tensorboardx

  #39741 commented on Jun 16, 2025 • 0 new comments

• [Core] Upgrading grpc to 1.57.0 causes perf regressions

  #39679 commented on Jun 16, 2025 • 0 new comments

• ray failed to register worker when I used vllm

  #39618 commented on Jun 16, 2025 • 0 new comments

• [rllib] Action space MultiDiscrete([11 5 1 2]) is not supported for DQN

  #39571 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Support JAX-(numpy)-based envs.

  #39528 commented on Jun 16, 2025 • 0 new comments

• [RLlib] Ray RLLib Dependencies Version Information

  #39405 commented on Jun 16, 2025 • 0 new comments

• [RLlib] dreamerv3 causes debug code to be executed when running tune

  #39302 commented on Jun 16, 2025 • 0 new comments

• [Core] CPP Interface crashes on Ray.Init()

  #39252 commented on Jun 16, 2025 • 0 new comments

• ValueError: Must set agent_id on policy config

  #39246 commented on Jun 16, 2025 • 0 new comments

• [Core] Actor retry count is consumed because the task is retried when actor is still alive.

  #39110 commented on Jun 16, 2025 • 0 new comments

• [Core] Memory Leak

  #38877 commented on Jun 16, 2025 • 0 new comments

• [Tune] Leaky core concepts in Ray Tune documentation

  #38781 commented on Jun 16, 2025 • 0 new comments

• [GCS]Support different backend for GCS instead of Redis

  #10356 commented on Jun 16, 2025 • 0 new comments

• [metrics] Better way of grouping metric definitions

  #10341 commented on Jun 16, 2025 • 0 new comments

• [Core] WorkerThreadContext semantics are incorrect for async Python actors.

  #10324 commented on Jun 16, 2025 • 0 new comments

• [ray] Programatically expose the amount of memory available in the object store

  #10278 commented on Jun 16, 2025 • 0 new comments

• [tune] Improve the serialization diagnoser by providing deeper introspection

  #10263 commented on Jun 16, 2025 • 0 new comments

• [tune] Usability issues

  #10248 commented on Jun 16, 2025 • 0 new comments

• [ray] Support mypy

  #10244 commented on Jun 16, 2025 • 0 new comments

• Removed the following hyperparameter values when logging to tensorboard: ... [tune]

  #10166 commented on Jun 16, 2025 • 0 new comments

• [dask-on-ray] ValueError on read-only memory

  #10124 commented on Jun 16, 2025 • 0 new comments

• [cli/docs] Provide example commands in the CLI docstrings.

  #10079 commented on Jun 16, 2025 • 0 new comments

• [Placement Group] Placement group dashboard

  #9775 commented on Jun 16, 2025 • 0 new comments

• Ray issue with serializing pytorch objects only when running on 40+ cores

  #9752 commented on Jun 16, 2025 • 0 new comments

• Ray typing IDE code completion support

  #9623 commented on Jun 16, 2025 • 0 new comments

• [Cluster][Task Schedule] Remote function is not executing without any errors

  #9598 commented on Jun 16, 2025 • 0 new comments

• [core] RayConfig does not get set properly after multiple `ray.init` calls

  #9545 commented on Jun 16, 2025 • 0 new comments

• [New scheduler] Performance optimization

  #9487 commented on Jun 16, 2025 • 0 new comments

• [New scheduler] Release testing

  #9486 commented on Jun 16, 2025 • 0 new comments

• Specify network interface to use / RuntimeError: Redis has started but no raylets have registered yet.

  #9456 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] Autoscaler prints (harmless) errors every 30 mins and then kills the workers in GCP cluster

  #9368 commented on Jun 16, 2025 • 0 new comments

• [Core] Core Worker Actor Handle GC.

  #9342 commented on Jun 16, 2025 • 0 new comments

• Graph related applications

  #9324 commented on Jun 16, 2025 • 0 new comments

• Options Support for Actor Methods

  #9296 commented on Jun 16, 2025 • 0 new comments

• `ray stop` should not kill all redis-server processes

  #11513 commented on Jun 16, 2025 • 0 new comments

• [core] Track the number of connection and use shared pool whenever possible for grpc clients.

  #11445 commented on Jun 16, 2025 • 0 new comments

• ray commandline tools raise exceptions if you forget the YAML config file

  #11396 commented on Jun 16, 2025 • 0 new comments

• [Autoscaler] Placement group rescheduling over-allocates resources

  #11372 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] request_resources with partial instance availability leads to workers never shutting down

  #11367 commented on Jun 16, 2025 • 0 new comments

• how to add two-timescales Learning rate schedule in coustom policy?

  #11328 commented on Jun 16, 2025 • 0 new comments

• [Autoscaler] Add additional gpu types to util.accelerators

  #11160 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] Worker node container is not removed after ray down?

  #11098 commented on Jun 16, 2025 • 0 new comments

• `ray stop` should wait for processes to exit

  #10955 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] node type preferences

  #10929 commented on Jun 16, 2025 • 0 new comments

• Private/onprem clusters always need explicit ssh_private_key in docker

  #10838 commented on Jun 16, 2025 • 0 new comments

• [docs] Add examples for using custom resources

  #10808 commented on Jun 16, 2025 • 0 new comments

• Autoscaler should set RAY_ADDRESS environment variable

  #10752 commented on Jun 16, 2025 • 0 new comments

• Stop using `file_mounts` for ray_bootstrap_config & ray_bootstrap_key

  #10743 commented on Jun 16, 2025 • 0 new comments

• [Java] Remove Java 9/10/11 warnings

  #10673 commented on Jun 16, 2025 • 0 new comments

• [Documentation] need for default_resource_requests when using custom train function

  #10572 commented on Jun 16, 2025 • 0 new comments

• [rllib] action from policy with Tuple action space has wrong shape

  #10516 commented on Jun 16, 2025 • 0 new comments

• [tune] String summarization/representations for user objects

  #10489 commented on Jun 16, 2025 • 0 new comments

• [tune] Add regression test for avoiding extraneous output

  #10485 commented on Jun 16, 2025 • 0 new comments

• [GCS]Remove tightly coupled Redis code path from Python

  #10359 commented on Jun 16, 2025 • 0 new comments

• [GCS]Support Sharding GCS server

  #10358 commented on Jun 16, 2025 • 0 new comments

• [GCS]Support Multi-threaded GCS server.

  #10357 commented on Jun 16, 2025 • 0 new comments

• [ray] constant memory usage increase of actor using actor handle.

  #9232 commented on Jun 16, 2025 • 0 new comments

• Code coverage tracker

  #5473 commented on Jun 16, 2025 • 0 new comments

• [ray] ray misuse gpu in docker container

  #5245 commented on Jun 16, 2025 • 0 new comments

• On a background thread, `ray.wait` doesn't timeout until another method on the actor is called

  #4934 commented on Jun 16, 2025 • 0 new comments

• Ray is not propagating variable types correctly

  #4463 commented on Jun 16, 2025 • 0 new comments

• [tune] Support nesting grid_search in lambdas

  #3466 commented on Jun 16, 2025 • 0 new comments

• Retry policy when a worker crashes: a hook missing?

  #2635 commented on Jun 16, 2025 • 0 new comments

• Task introspection

  #2617 commented on Jun 16, 2025 • 0 new comments

• ray start does not restart failed processes

  #2587 commented on Jun 16, 2025 • 0 new comments

• CI test linux://rllib:learning_tests_cartpole_dqn_gpu is flaky

  #46683 commented on Jun 16, 2025 • 0 new comments

• [rllib] flattening error in gym.spaces.Sequence

  #45563 commented on Jun 16, 2025 • 0 new comments

• cannot import name 'EPISODE_RETURN_MEAN' from 'ray.rllib.utils.metrics'

  #45453 commented on Jun 16, 2025 • 0 new comments

• error: No such option: --torch

  #45452 commented on Jun 16, 2025 • 0 new comments

• [Core] Unable to run worker with virtual environment without installing dashboard

  #45410 commented on Jun 16, 2025 • 0 new comments

• [RLlib] How to support gymnasium graph obs space?

  #45290 commented on Jun 16, 2025 • 0 new comments

• Ray Cluster does not work across multiple docker containers

  #45252 commented on Jun 16, 2025 • 0 new comments

• [Core] Worker crashes unexpectedly due to frequent triggering of OOM

  #45244 commented on Jun 16, 2025 • 0 new comments

• Ray Cluster: Failed to create a ray cluster using running container

  #45148 commented on Jun 16, 2025 • 0 new comments

• [Rllib] Rllib provides wrong state batch size during "bug check" batches on torch custom model

  #45131 commented on Jun 16, 2025 • 0 new comments

• [RLlib] ValueError in initialization of ImpalaTF2Policy

  #45050 commented on Jun 16, 2025 • 0 new comments

• [core] GcsSubscriber hangs in shutdown if the connection broke on MacOS

  #45044 commented on Jun 16, 2025 • 0 new comments

• Workflow: Reading workflow status can lead to corrupted json reads.

  #45027 commented on Jun 16, 2025 • 0 new comments

• [Core] `ray.wait` not actually wait until ready when the task is longer than 12 days

  #44909 commented on Jun 16, 2025 • 0 new comments

• Invalid memory access in RedisAsioClient/RedisAsyncContext on shutdown

  #9074 commented on Jun 16, 2025 • 0 new comments

• Performance issue with many large tasks on 10 node cluster.

  #8950 commented on Jun 16, 2025 • 0 new comments

• Ray Dashboard Head-node CLI [autoscaler]

  #8450 commented on Jun 16, 2025 • 0 new comments

• Support TPUs across all of Ray

  #8260 commented on Jun 16, 2025 • 0 new comments

• [core] ray.init does not work if run in a node with external ip while the cluster is started internally

  #8244 commented on Jun 16, 2025 • 0 new comments

• [docs][autoscaler] additional dependencies needs to be mentioned to build your own autoscaler image

  #8235 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] Api instead of CLI to interact with cluster.

  #8036 commented on Jun 16, 2025 • 0 new comments

• Incorrect unreconstructable error message and raise different exception.

  #7804 commented on Jun 16, 2025 • 0 new comments

• ray.wait hangs with no warning or error when local object store is too small to receive object

  #7802 commented on Jun 16, 2025 • 0 new comments

• Segmentation Fault when using multiprocessing.Queue

  #7793 commented on Jun 16, 2025 • 0 new comments

• ray.wait with local_mode=True blocks for a very long time

  #7741 commented on Jun 16, 2025 • 0 new comments

• Awesome: algorithm selection helper & diagrams

  #7722 commented on Jun 16, 2025 • 0 new comments

• Ray hangs when machine is disconnected from network

  #7696 commented on Jun 16, 2025 • 0 new comments

• [docs] Clarify that in K8s the jobs need to be launched from the workers

  #7188 commented on Jun 16, 2025 • 0 new comments

• [ray] tasks running in docker containers are not stopped on local cluster

  #6898 commented on Jun 16, 2025 • 0 new comments

• [dist] Release notes for Java And other Languages

  #6608 commented on Jun 16, 2025 • 0 new comments

• Ray.wait causes node to hang if there are too many object ids

  #6403 commented on Jun 16, 2025 • 0 new comments

• Performance issues with defining remote functions and actor classes from within tasks.

  #6240 commented on Jun 16, 2025 • 0 new comments

• TypeError: can't pickle CudnnModule objects

  #5947 commented on Jun 16, 2025 • 0 new comments

• Profiling ray tasks includes ray initialization time

  #5832 commented on Jun 16, 2025 • 0 new comments

• Make it possible to see resource deadlocks through web UI.

  #5789 commented on Jun 16, 2025 • 0 new comments

• [autoscaler] Raise better error message if `ssh_user` is not correct

  #5772 commented on Jun 16, 2025 • 0 new comments