Insights: ray-project/ray

1 Release published by 1 person

• ray-2.3.0 Ray-2.3.0

  published Feb 24, 2023

66 Pull requests merged by 34 people

• Fixed dashboard test due to bad dict.pop call

  #32559 merged Feb 25, 2023

• [core] Update OpenCensus

  #32553 merged Feb 25, 2023

• [data] [streaming] Fix deque race condition and increase max tasks in flight per actor

  #32800 merged Feb 25, 2023

• Remove usage_lib.LibUsageRecorder

  #32806 merged Feb 25, 2023

• [Doc] Add Configuring Batch Predictor section in Predictors User Guides

  #32436 merged Feb 25, 2023

• [Doc|AIR] Add Pytorch ResNet batch prediction example

  #32470 merged Feb 25, 2023

• [Serve] Add hash function of RunningReplicaInfo

  #32772 merged Feb 24, 2023

• Make Write op extend AbstractMap operator

  #32538 merged Feb 24, 2023

• [Datasets] Fix `from_items` parallelism to create the expected number of blocks.

  #32821 merged Feb 24, 2023

• [core] Fix the race condition in the new resource broadcasting.

  #32798 merged Feb 24, 2023

• [Datasets] Support different number of blocks/rows per block in zip().

  #32795 merged Feb 24, 2023

• Don't call ray.init() when num_workers is zero

  #32820 merged Feb 24, 2023

• Modify the 'MemoryTracingCallbacks' to improve comparability of metrics and reduce noise in memory allocations

  #31859 merged Feb 24, 2023

• [RLlib] IMPALA trainer tf

  #32713 merged Feb 24, 2023

• [RLlib] Fixed telemetry RLlib bug

  #32799 merged Feb 24, 2023

• [Serve] Delete apps not listed in config

  #32665 merged Feb 24, 2023

• Fix or re-revert https://github.com/ray-project/ray/pull/32690

  #32807 merged Feb 24, 2023

• [ci] Install the dependence first in CI images.

  #32797 merged Feb 24, 2023

• Improve title for SEO

  #32792 merged Feb 24, 2023

• [Release] Unpin anyscale version in the OSS release tests

  #32784 merged Feb 23, 2023

• [Datasets] [Operator Fusion - 2/N] Data layer performance/bug fixes and tweaks.

  #32744 merged Feb 23, 2023

• [Jobs] Add end to end test for multiple drivers in a Ray Job

  #32485 merged Feb 23, 2023

• Fix disk metric showing double the actual value

  #32674 merged Feb 23, 2023

• [build_base] Change ordering of installing dependence and build

  #32780 merged Feb 23, 2023

• [Tune] Sync trial artifacts to cloud

  #32334 merged Feb 23, 2023

• [autoscaler][CLI logger] Stop overriding non-default record factories.

  #32678 merged Feb 23, 2023

• [core] Allow worker Python interpeters to inherit -S and -s flags

  #32690 merged Feb 23, 2023

• [docs] bugfix: make secondary nav items clickable again

  #32695 merged Feb 23, 2023

• [AIR] Rewording in `analyze_tuning_results.ipynb`

  #32671 merged Feb 23, 2023

• [core][state] Task backend - disable verbose print.

  #32764 merged Feb 23, 2023

• [data] Streaming executor fixes #2

  #32759 merged Feb 23, 2023

• Prepend task id if it exists to "ray task list --detail"

  #32353 merged Feb 23, 2023

• [Serve] Fix serve.shutdown() to delete deployments from all applications

  #32631 merged Feb 23, 2023

• [core]Update boost in ray deps

  #32550 merged Feb 22, 2023

• [release][train] Add `FileLock` to `tensorflow_mnist_example`.

  #32712 merged Feb 22, 2023

• [Release] Fix horovod release test tags

  #32743 merged Feb 22, 2023

• [Autoscaler] Bump ebs volume size in default configs

  #32708 merged Feb 22, 2023

• [data] [streaming] Limit operator memory by downstream usage only

  #32673 merged Feb 22, 2023

• Split an AbstractUDFMap out of AbstractMap

  #32573 merged Feb 22, 2023

• [Rllib] Enable tf eager tracing on the training side

  #32724 merged Feb 22, 2023

• [RLlib] CNN Models for RLModules

  #32515 merged Feb 22, 2023

• [3/n] Stabilize GCS/Autoscaler interface: Expose task/actor resource demands and node resource usage

  #32561 merged Feb 22, 2023

• [Release Tests] Properly use `ttl-hours` tag

  #32727 merged Feb 22, 2023

• [CI] Add streaming executor CI

  #32643 merged Feb 22, 2023

• Streaming executor fixes

  #32658 merged Feb 22, 2023

• [Core][C++ Worker]Optimize the log printing and code of the LD_LIBRARY_PATH environment variable

  #32682 merged Feb 22, 2023

• Revert "[AIR/Data] Make `BatchPredictor` lazy"

  #32710 merged Feb 22, 2023

• [Data] Fix legacy do_write

  #32661 merged Feb 21, 2023

• [Doc] Clarify that object_spilling_config only needs to be specified on the head node

  #32702 merged Feb 21, 2023

• [Release] no-terminate workaround for anyscale job

  #32705 merged Feb 21, 2023

• [Doc] Improve docstring of ray.remote parameters

  #32701 merged Feb 21, 2023

• [docs] Improve warning around Ray Debugger

  #32703 merged Feb 21, 2023

• [Datasets] Add `read_tfrecords` comment about reading compressed files

  #32544 merged Feb 21, 2023

• [runtime_env] Print correct default for expiration_s environment variable

  #32230 merged Feb 21, 2023

• [Jobs] Add integration test for JobInfo GCS endpoint

  #32186 merged Feb 21, 2023

• [Doc] [runtime env] Add runtime env architecture doc

  #32502 merged Feb 21, 2023

• [Core] Expose KV MultiGet rpc to python GCS client

  #32600 merged Feb 21, 2023

• [AIR/Data] Add `collate_fn` to `iter_torch_batches`

  #32412 merged Feb 21, 2023

• [AIR/Data] Make `BatchPredictor` lazy

  #32510 merged Feb 21, 2023

• [Release Tests] Update tags to conform to new requirements

  #32689 merged Feb 21, 2023

• [air] Fix NoneType error loading TorchCheckpoint through `from_uri`.

  #32386 merged Feb 21, 2023

• [core] Fix error for connecting to existing cluster from OSX/Windows driver

  #32646 merged Feb 21, 2023

• [Core][C++ Worker]Add ray_job_namespace param for c++ worker driver

  #32681 merged Feb 21, 2023

• Added a blog example for MMT in the use cases gallery

  #32647 merged Feb 20, 2023

• [Core][C++ Worker] libray_api.so exports absl::flag symbols

  #32648 merged Feb 20, 2023

• [Release] Update paging Anyscale API call

  #32679 merged Feb 20, 2023

54 Pull requests opened by 35 people

• [RLLib] Fix Dirichlet `deterministic_sample` and add `TorchDirichlet` to `catalog`

  #32684 opened Feb 20, 2023

• [Core] Disable debug dump

  #32685 opened Feb 20, 2023

• [core] additional telemetry for worker pool

  #32686 opened Feb 20, 2023

• [Core][C++ Worker]Add metric API for c++ worker to align the metric API of java and python

  #32692 opened Feb 21, 2023

• [Dataset] Correct `Dataset.to_pandas()` conversion for dicts containing ndarrays

  #32709 opened Feb 21, 2023

• [Datasets] Truncate lines from `Dataset.__repr__` involving long column names

  #32722 opened Feb 21, 2023

• [RLlib] RNNs and RLModules

  #32723 opened Feb 21, 2023

• [core][state] Task backend - Fix actor creation states when failed

  #32726 opened Feb 22, 2023

• [core][benchmark] minor clean up - simplify results reporter for benchmark tests

  #32728 opened Feb 22, 2023

• [Docs] Update on-premises.md for using clusters under windows.

  #32731 opened Feb 22, 2023

• fix docs: update broken links and edit for conciseness

  #32733 opened Feb 22, 2023

• feat: use `jupyterlab` instead of `jupyter` PyPI package

  #32734 opened Feb 22, 2023

• [Doc] Make front page images non clickable

  #32738 opened Feb 22, 2023

• [ci] Allow using local wheels to run release tests.

  #32739 opened Feb 22, 2023

• Global logging format changes

  #32741 opened Feb 22, 2023

• [Datasets] [Operator Fusion - 3/N] Add operator fusion benchmark.

  #32747 opened Feb 22, 2023

• [WIP] combine workloads and use cases pages

  #32748 opened Feb 22, 2023

• [Datasets] [Operator Fusion - 4/N] Add metrics collection to data layer.

  #32749 opened Feb 22, 2023

• [serve] Update Gradio example in docs

  #32754 opened Feb 22, 2023

• Grafana default head node ip

  #32758 opened Feb 22, 2023

• [WIP] [Data] Allow generator UDFs for `map_batches`

  #32767 opened Feb 23, 2023

• [WIP; RLlib] Checkpoints are convertible to msgpack (to make them python-version-independent).

  #32779 opened Feb 23, 2023

• Bump markdown-it-py from 1.1.0 to 2.2.0 in /python

  #32785 opened Feb 23, 2023

• [Serve] Validate all applications have unique names and route prefixes in ServeDeploySchema

  #32787 opened Feb 23, 2023

• Bump gradio from 3.11 to 3.13.2 in /python

  #32789 opened Feb 23, 2023

• Improve title for SEO

  #32790 opened Feb 23, 2023

• Improve SEO with title change

  #32791 opened Feb 23, 2023

• Revert "Revert "[AIR/Data] Make `BatchPredictor` lazy""

  #32796 opened Feb 24, 2023

• [minor] fix dl documentation

  #32801 opened Feb 24, 2023

• [Doc][Anyscale] Workspace template examples

  #32802 opened Feb 24, 2023

• Revert "[core] Add opt-in flag for Windows and OSX clusters, update r…

  #32805 opened Feb 24, 2023

• [RLlib] [DO NOT MERGE] Enable ppo learner

  #32808 opened Feb 24, 2023

• [Datasets] Make binary data source output Arrow block

  #32809 opened Feb 24, 2023

• [RLlib] Introduce composite models to

  #32811 opened Feb 24, 2023

• Air/lightning trainer

  #32813 opened Feb 24, 2023

• [docs][examples] Ray Core walkthrough for beginners

  #32814 opened Feb 24, 2023

• [docs][examples] simple MR with Ray Core

  #32816 opened Feb 24, 2023

• [RLlib] Example for how to use custom GNN models based on TF-GNN

  #32818 opened Feb 24, 2023

• [RLlib] Initialize the default view requirements when using the learner group api

  #32825 opened Feb 24, 2023

• Remove use of ATOMIC_VAR_INIT

  #32828 opened Feb 24, 2023

• Deflake actor_checkpointing.py

  #32829 opened Feb 24, 2023

• Update Get Started link; add Anyscale logo to the button.

  #32830 opened Feb 24, 2023

• [K8S] Remove port limitations in the K8S networkpolicies

  #32831 opened Feb 24, 2023

• [RLlib] Fix repro ppo learner

  #32832 opened Feb 24, 2023

• [Release] Adding various missing wheels to download_wheels.

  #32833 opened Feb 24, 2023

• [core][state] Adding task failure type to state API

  #32834 opened Feb 24, 2023

• [Datasets] Streaming executor fixes #3

  #32836 opened Feb 25, 2023

• [Test] Migrate chaos_dataset_shuffle_random_shuffle_1tb and chaos_dataset_shuffle_push_based_random_shuffle_1tb tests to v2 stack

  #32838 opened Feb 25, 2023

• [Serve] Add http request latency

  #32839 opened Feb 25, 2023

• [core] Update protobuf

  #32840 opened Feb 25, 2023

• [WIP] Guard against google colab environments for ipywidgets and html reprs

  #32841 opened Feb 25, 2023

• [doc][xlang] Update xlang doc

  #32844 opened Feb 25, 2023

• [data](deps): Bump dask[complete] from 2022.10.1 to 2023.2.1 in /python/requirements/data_processing

  #32845 opened Feb 25, 2023

• [RLlib] Improve setting rollout_fragment_length logic to get the expected train_batch_size

  #32846 opened Feb 25, 2023

45 Issues closed by 25 people

• [Dashboard] Failing test due to bad call to `dict.pop`

  #32557 closed Feb 25, 2023

• [data] [streaming] deque mutated during iteration

  #32765 closed Feb 25, 2023

• [data] [streaming] Increase default max_tasks_in_flight_per_actor to 4

  #32770 closed Feb 25, 2023

• [Ray FT] Ray 2.3.0 cannot connect to Redis

  #32794 closed Feb 25, 2023

• [Serve] max_concurrent_queries=1 is ignored when autoscaling

  #32652 closed Feb 24, 2023

• [docs] Improving the quality of search

  #31094 closed Feb 24, 2023

• [CI] `windows://python/ray/tests:test_runtime_env_working_dir_remote_uri` is failing/flaky on master.

  #29001 closed Feb 24, 2023

• [CI] `windows://python/ray/tests:test_runtime_env_working_dir_2` is failing/flaky on master.

  #29000 closed Feb 24, 2023

• [Datasets] Improve user experience of `zip()`

  #32375 closed Feb 24, 2023

• [core] dump failure reason for task attempts on oom failure

  #28927 closed Feb 24, 2023

• [Serve] Applications are not deleted if removed from config file

  #32664 closed Feb 24, 2023

• [tune] Sync artifacts in trial/ directories to cloud

  #30071 closed Feb 23, 2023

• [autoscaler] _patched_makeRecord needs to make use of logging module's global _logRecordFactory

  #13795 closed Feb 23, 2023

• [core][Bug] Ray processes escaping hermetic python environment

  #22977 closed Feb 23, 2023

• [RLlib] ZeroDivisionError: integer division or modulo by zero because "batch size" is 0.

  #30881 closed Feb 23, 2023

• [Core] API for getting the size of an object

  #32768 closed Feb 23, 2023

• [Data] Possible race condition in Streaming executor

  #32766 closed Feb 23, 2023

• [Core / State observability] Keep the task state after it is finished.

  #24961 closed Feb 23, 2023

• [RLlib; Pre-checks/better failure behavior] Add flags for env checks

  #20582 closed Feb 23, 2023

• [RLlib; Pre-checks/better failure behavior] Multi-agent environment checks

  #20580 closed Feb 23, 2023

• [RLlib; Pre-checks/better failure behavior] Merge env checks onto critical code path

  #20581 closed Feb 23, 2023

• [RLlib; Pre-checks/better failure behavior] Base Env Checks

  #20579 closed Feb 23, 2023

• [Bug] [RLlib] module 'gym.wrappers' has no attribute 'Monitor'

  #23286 closed Feb 23, 2023

• [RLlib] pip install ray[rllib]==2.0.1 fails due to missing [rllib] extra

  #29652 closed Feb 23, 2023

• [Ray 2.3 release] `rllib_learning_tests_leela_chess_zero` fails with `AttributeError`

  #32147 closed Feb 23, 2023

• [Serve] `serve.shutdown()` only deletes deployments under app `""`

  #32623 closed Feb 23, 2023

• Observability: provide visibility into retried attempts

  #26524 closed Feb 22, 2023

• Memory management <Ray Core>

  #32706 closed Feb 22, 2023

• [Cluster VM launcher] Small ebs size in yaml file makes the example fail

  #32707 closed Feb 22, 2023

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

  #32430 closed Feb 22, 2023

• [Core] ray installed from conda does not include the dashboard

  #30834 closed Feb 22, 2023

• [Release tests] [rllib] `rllib_learning_tests_impala_tf` release test flaky

  #31885 closed Feb 22, 2023

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

  #31970 closed Feb 21, 2023

• [Tune] Document the parameter passing and connection from param_space to trainable

  #30158 closed Feb 21, 2023

• [tune] Control usage of resource

  #28161 closed Feb 21, 2023

• [Datasets] Support reading compressed TFRecord file

  #32205 closed Feb 21, 2023

• [Core] Expose MultiGet Internal KV operation in Python

  #32122 closed Feb 21, 2023

• [Ray Data] Introduce collate_fn argument in Dataset.iter_torch_batches

  #32224 closed Feb 21, 2023

• Torch batch prediction fails

  #32284 closed Feb 21, 2023

• [Core] regressions in ray start --head usage in OSX

  #32640 closed Feb 21, 2023

• Long-running redis `KEYS` command caused external MemoryDB storage to hang

  #32537 closed Feb 20, 2023

• [release][core] chaos_many_actors is failing due to ray.wait regression

  #30694 closed Feb 20, 2023

• [Core][nightly-test] chaos_dataset_shuffle_sort_1tb failing due to OOM killer

  #30680 closed Feb 20, 2023

• [placement groups] Pending tasks not removed after job terminated

  #15778 closed Feb 20, 2023

• [tune/Autoscaler] Autoscaler made demands on a node while knowing this node has died / Cluster hangs randomly when a spot instance worker node is killed.

  #32579 closed Feb 19, 2023

68 Issues opened by 40 people

• [Ray Core] Ray Dashboard Actor stack trace breaks on Windows

  #32847 opened Feb 25, 2023

• [Release][AIR] Inconsistent batch prediction performance is causing `air_benchmark_xgboost` to be flaky

  #32843 opened Feb 25, 2023

• [Release][Tune] `tune_cloud_aws_durable_upload` failed after restoring experiment

  #32842 opened Feb 25, 2023

• [Ray: Serve] Model Composition primitives should be part of Serve Core API docs.

  #32837 opened Feb 25, 2023

• [core][state] Children tasks shouldn't be marked as failed when parent raises application errors

  #32835 opened Feb 25, 2023

• [docs] Airflow Integration example

  #32827 opened Feb 24, 2023

• [core][state] Task backend : already submitted cancelled task showing up as finished

  #32826 opened Feb 24, 2023

• [Doc] Make docstring signatures more readable

  #32824 opened Feb 24, 2023

• [AIR][Tune] Make trial checkpoint + artifact upload happen atomically

  #32823 opened Feb 24, 2023

• [core][oom] large scale workloads reporting OOM deadlock

  #32822 opened Feb 24, 2023

• [Core][API] get error from ObjectRef without the overhead of fetching the actual data

  #32817 opened Feb 24, 2023

• [RLlib] Training iteration timer computes misleading timing

  #32815 opened Feb 24, 2023

• [RLlib] A3C has problems with the horizon option removed

  #32812 opened Feb 24, 2023

• [Tune] During multi-GPU training (using mp.spawn), ray.tune.report does not take effect.

  #32810 opened Feb 24, 2023

• [CI] `windows://python/ray/tests:test_basic_5_client_mode` is failing/flaky on master.

  #32804 opened Feb 24, 2023

• [core][agent] Dashboard agent emits deprecation warning from ariogrpc's calling get_event_loop

  #32793 opened Feb 23, 2023

• Optimizing SEO performance

  #32788 opened Feb 23, 2023

• [3/11, 3/12 - SF] Join us in Hacker Dojo for a mix of Ray training and networking!

  #32786 opened Feb 23, 2023

• [AIR][Tune] Make `SyncConfig(sync_artifacts=False)` work fully (for all syncing methods)

  #32783 opened Feb 23, 2023

• [AIR][Tune] Enable trial artifacts to be restored from cloud and updated correctly

  #32782 opened Feb 23, 2023

• [air] survey all air created tmp folders, prefix them with a unique identifier

  #32781 opened Feb 23, 2023

• Issue on page /index.html /How to make MIG GIs visible in a Ray session, for allocating a different worker in a different MIG GI.

  #32778 opened Feb 23, 2023

• [dataset] Support for data loading/writing to BigQuery

  #32777 opened Feb 23, 2023

• [Core] Permission denied error when using tune.Tuner() on a shared machine

  #32776 opened Feb 23, 2023

• [air] lightning integration

  #32775 opened Feb 23, 2023

• [CI]Mac OS Pipeline not kicked off

  #32774 opened Feb 23, 2023

• [AIR: Train] DatasetConfig "use_stream_api" option raises RuntimeError

  #32773 opened Feb 23, 2023

• [AIR] Set informative name for Trainers/Trainables

  #32763 opened Feb 23, 2023

• [core][state] ray terminate showing up in task api

  #32762 opened Feb 23, 2023

• [Dataset] Make `limit()` lazy and add logical operator for it

  #32761 opened Feb 23, 2023

• [Tune] failure when using more than one GPU

  #32760 opened Feb 22, 2023

• [data] [streaming] Add Jupyter progress bar

  #32757 opened Feb 22, 2023

• [Dataset] Add optional schema param to read_tfrecords()

  #32756 opened Feb 22, 2023

• [Dataset] Set informative Dataset task/actor name to propagate to Ray dashboard

  #32753 opened Feb 22, 2023

• [Ray : Serve] The Gradio Ray Serve tutorial does not work in Ray 2.2 and the master

  #32752 opened Feb 22, 2023

• [RLLIB] restore_1_of_n_agents_from_checkpoint does not actually seem to work

  #32751 opened Feb 22, 2023

• [release] Regression of actor creation benchmark

  #32750 opened Feb 22, 2023

• [data] Block splitting doesn't work for single large tensor outputs of map_batches()

  #32746 opened Feb 22, 2023

• [data] [streaming] Support autoscaling with Data streaming execution

  #32745 opened Feb 22, 2023

• [core] long running node failure when light weight resource broadcasting is on

  #32742 opened Feb 22, 2023

• [HuggingFace] Fail transformers fine-tuning tutorial when running with Ray Client

  #32740 opened Feb 22, 2023

• [Doc] Front page big buttons link to images instead of docs pages

  #32737 opened Feb 22, 2023

• [Jobs] Specifying API server address without `http://` raises unhelpful error "Unable to connect to GCS at localhost:8265"

  #32736 opened Feb 22, 2023

• [<Ray component: Core|RLlib|etc...>]

  #32735 opened Feb 22, 2023

• [AIR][Train] Multilabel classification with `SklearnTrainer`

  #32732 opened Feb 22, 2023

• [nightly] stress_test_state_api_scale failed

  #32730 opened Feb 22, 2023

• ray error packaging zip file even with smart_open installed

  #32729 opened Feb 22, 2023

• [Train] TorchTrainer does not free all GPUs on shutdown

  #32725 opened Feb 22, 2023

• [Release][Core] Migrate `aws_cluster_launcher` test to v2 stack and Anyscale Jobs

  #32721 opened Feb 21, 2023

• [Release][Serve] Migrate `k8s_serve_ha_test` tests to v2 stack and Anyscale Jobs

  #32720 opened Feb 21, 2023

• [Release][Data] Migrate `chaos_dataset_shuffle_random_shuffle_1tb` and `chaos_dataset_shuffle_push_based_random_shuffle_1tb` tests to v2 stack and Anyscale Jobs

  #32719 opened Feb 21, 2023

• [Release][Data] Migrate `dataset_shuffle_random_shuffle_1tb` test to v2 stack and Anyscale Jobs

  #32718 opened Feb 21, 2023

• [Release][Serve] Migrate `long_running_serve_failure` test to v2 stack and Anyscale Jobs

  #32717 opened Feb 21, 2023

• [Release][Tune] Migrate `tune_scalability_result_throughput_single_node` tests to v2 stack and Anyscale Jobs

  #32716 opened Feb 21, 2023

• [Release][Tune] Migrate `tune_cloud_gcp_k8s_*` tests to v2 stack and Anyscale Jobs

  #32715 opened Feb 21, 2023

• [Ray Dataset] Streaming executor trying to consume more resources than limit.

  #32714 opened Feb 21, 2023

• [Serve] No http latency metrics

  #32711 opened Feb 21, 2023

• [Runtime Env] Add docstring for public class methods and attributes

  #32704 opened Feb 21, 2023

• [AIR] Add an example of using remote storage (s3) with endpoint url

  #32700 opened Feb 21, 2023

• [Ray Workflow] example of time-series dependent tasks (w/ different resolution)

  #32699 opened Feb 21, 2023

• [tune] Add suggestions on when `reuse_actor` should be set to false.

  #32698 opened Feb 21, 2023

• [RLlib] check for gym env should assume `issubclass` instead of `isinstance`

  #32697 opened Feb 21, 2023

• Incorrect torch.exp in autorgressive action torch example

  #32696 opened Feb 21, 2023

• [Core][Object Store] Object Store to manage files in the cluster

  #32694 opened Feb 21, 2023

• [Windows][Installation] Install of Ray using Windows subsystem for Linux (WSL) fails to start the Ray dashboard

  #32693 opened Feb 21, 2023

• [Core][C++ Worker]Add metric API for c++ worker to align the metric API of java and python

  #32691 opened Feb 21, 2023

• [Core][C++ Worker]Optimize the log printing and code of the LD_LIBRARY_PATH environment variable

  #32683 opened Feb 20, 2023

• [Core] [C++ Worker]Add ray_job_namespace param for c++ worker driver

  #32680 opened Feb 20, 2023

215 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.

• [RLlib] Introduce Catalog skeleton for RLModules

  #32069 commented on Feb 24, 2023 • 80 new comments

• [Doc|AIR] Add Pytorch ResNet fine-tuning example

  #32226 commented on Feb 23, 2023 • 16 new comments

• [Core] Job gets stuck when submit multiple jobs to the Ray cluster

  #32215 commented on Feb 23, 2023 • 11 new comments

• Use PID to assign rainbow colors to console output prefixes.

  #26900 commented on Feb 25, 2023 • 9 new comments

• [Serve] Add Multi-App REST API

  #32642 commented on Feb 25, 2023 • 9 new comments

• [Train] Better Hugging Face Accelerate Integration

  #32572 commented on Feb 21, 2023 • 8 new comments

• [Datasets] Improved naming of Ray Data `map` tasks

  #32585 commented on Feb 24, 2023 • 8 new comments

• Incomplete specification of necessary gymnasium extras in `python/setup.py`?

  #31587 commented on Feb 22, 2023 • 7 new comments

• [Tune] Pause trials scheduled with the ASHA scheduler when they reach a milestone

  #32634 commented on Feb 24, 2023 • 7 new comments

• [serve]add java serve http ingress with jersey application handler

  #31322 commented on Feb 22, 2023 • 6 new comments

• [Serve]Make http retries tunable

  #32532 commented on Feb 23, 2023 • 6 new comments

• [core][state] Task backend - worker side optimization

  #32574 commented on Feb 23, 2023 • 6 new comments

• [Cluster Launcher] Add release test for minimal aws config from docs

  #32670 commented on Feb 25, 2023 • 5 new comments

• add link to user guide for Batch Predictor

  #32672 commented on Feb 22, 2023 • 5 new comments

• Failed to start the dashboard, return code 1

  #32312 commented on Feb 25, 2023 • 4 new comments

• [Ray API] io.ray.api.exception.RayActorException: The actor <id> died unexpectedly before finishing this task.

  #31857 commented on Feb 25, 2023 • 4 new comments

• Add resource limit for worker in container

  #16871 commented on Feb 25, 2023 • 4 new comments

• Optimize submit Ray task with many ObjectRef arguments

  #32478 commented on Feb 24, 2023 • 4 new comments

• [runtime env] Improve runtime env logs and remove doc for `SKIP_LOCAL_GC`

  #32663 commented on Feb 22, 2023 • 4 new comments

• [RFC] Async request support in Ray Serve

  #32292 commented on Feb 25, 2023 • 3 new comments

• [RLlib] Allow only vanilla replay buffer for qmix

  #32575 commented on Feb 23, 2023 • 3 new comments

• [core] Multi-process(?) / GPU processes do not seem to be freed after ctrl+c on cluster

  #31451 commented on Feb 22, 2023 • 2 new comments

• [Core] Moving "dead" worker files to `/tmp/ray/session_latest/logs/old` is confusing behavior for log shippers

  #32114 commented on Feb 22, 2023 • 2 new comments

• [Core] Apparent object store memory leak

  #25779 commented on Feb 23, 2023 • 2 new comments

• [Tune] time_attr paramter to async hyperband lacks meaningful description

  #32467 commented on Feb 24, 2023 • 2 new comments

• how to hide process command line parameters like --redis_password 5241590000000000 on command line gcs_server

  #29246 commented on Feb 25, 2023 • 2 new comments

• [Dashboard] Warning events' severity levels are "Info"

  #32012 commented on Feb 25, 2023 • 2 new comments

• [Core] Getting node id for usage in NodeAffinitySchedulingStrategy

  #28195 commented on Feb 25, 2023 • 2 new comments

• [Core][State Observability] Improve log tailing from log_client

  #28188 commented on Feb 25, 2023 • 2 new comments

• Import logging break ray

  #30443 commented on Feb 21, 2023 • 2 new comments

• [rllib] PPO, Actor-Critic, where are models?

  #7571 commented on Feb 19, 2023 • 1 new comment

• [Ray 2.3 Release] `shuffle_1tb_5000_partitions` failing with OOM-related errors

  #32272 commented on Feb 20, 2023 • 1 new comment

• [Core] Using multiprocessing to start a child process in a ray worker and starting a grandchild process in the child process again causes a hang

  #31263 commented on Feb 20, 2023 • 1 new comment

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

  #29420 commented on Feb 20, 2023 • 1 new comment

• [core][perf] single_node_2 large_object_time latency benchmark spikes

  #30711 commented on Feb 20, 2023 • 1 new comment

• [installation] Ray Installation broken with circular import on virtualenvs

  #29587 commented on Feb 20, 2023 • 1 new comment

• [core] `ray.get_runtime_context.get_assigned_resources()` does not return raw resources in PG

  #28814 commented on Feb 21, 2023 • 1 new comment

• [Core] com_github_grpc_grpc complie error in gcc version 9.2.1

  #32565 commented on Feb 21, 2023 • 1 new comment

• [Dataset] `Dataset.to_pandas()` method incorrectly converts tables with a nested dict column

  #31860 commented on Feb 21, 2023 • 1 new comment

• Dreamer error: frame_skip unexpected

  #32508 commented on Feb 21, 2023 • 1 new comment

• [Datasets] Write out directory-partitioned datasets

  #24879 commented on Feb 21, 2023 • 1 new comment

• [Core] ray.init() hangs/fails after "Started a local Ray instance."

  #31897 commented on Feb 22, 2023 • 1 new comment

• [Bazel] Build fails with Bazel 6.0.0: Constraints from "@bazel_tools" removed

  #31504 commented on Feb 22, 2023 • 1 new comment

• [Java] ray_dist.jar not included in darwin arm64 whl files

  #31319 commented on Feb 22, 2023 • 1 new comment

• [Feature] Looking for a way to cancel ray serve task

  #22450 commented on Feb 22, 2023 • 1 new comment

• [Core] Worker reuse sometimes fails

  #30985 commented on Feb 22, 2023 • 1 new comment

• [Core] Task with a TF process is not releasing gpu resources. max_calls=0 is not resolving the issue

  #32210 commented on Feb 22, 2023 • 1 new comment

• [core] objects owned by worker exceeding its lifetime (or its job)

  #32418 commented on Feb 23, 2023 • 1 new comment

• [data] [streaming] Add execution options into public Dataset API

  #31797 commented on Feb 23, 2023 • 1 new comment

• [cloudpickle]Can't pickle local object 'AbstractProcess.fill_mode.<locals>.<lambda>'

  #32656 commented on Feb 24, 2023 • 1 new comment

• [RLlib] - Apex DQN unable to run on CartPole due to ReplayBuffer API change

  #32568 commented on Feb 24, 2023 • 1 new comment

• [gcs] Better error message when failing to connect to GCS external storage

  #32541 commented on Feb 24, 2023 • 1 new comment

• [Clusters] [KubeRay] problem with pending actors' pods in Kubernetes

  #32651 commented on Feb 24, 2023 • 1 new comment

• [CI] `windows://python/ray/tests:test_object_spilling_asan` is failing/flaky on master.

  #29021 commented on Feb 24, 2023 • 1 new comment

• [Bug] No Logging Options for Serve Deployment and/or Ray generated Pool Actors

  #22605 commented on Feb 25, 2023 • 1 new comment

• [Core] Fail more gracefully when the actor cannot be reached due to network related issue

  #29656 commented on Feb 25, 2023 • 1 new comment

• [WIP] [docs] add glossary with term references

  #29605 commented on Feb 20, 2023 • 1 new comment

• Make ray.get(timeout=0) to throw timeout error

  #30210 commented on Feb 22, 2023 • 1 new comment

• [all_tests][ci] Add 310 for test wheels on linux

  #31343 commented on Feb 22, 2023 • 1 new comment

• Ci/clean notebooks

  #31746 commented on Feb 19, 2023 • 1 new comment

• [core] Fix ray client test for generator tasks

  #31814 commented on Feb 20, 2023 • 1 new comment

• Deflake test_dataset.py: split tensor tests

  #32540 commented on Feb 24, 2023 • 1 new comment

• Update copy for metrics documentation

  #32614 commented on Feb 24, 2023 • 1 new comment

• [core]Turn on light weight resource broadcasting.

  #32625 commented on Feb 25, 2023 • 1 new comment

• [Datasets] Unskip Datasets memory release tests

  #24757 commented on Feb 20, 2023 • 0 new comments

• [Data] run small scale of dataset chaos integration test in CI

  #27257 commented on Feb 20, 2023 • 0 new comments

• [Data] ray.data ignores schema if block contains zero rows

  #28505 commented on Feb 20, 2023 • 0 new comments

• [Datasets] Add partitions per-block instead of per-batch

  #28867 commented on Feb 20, 2023 • 0 new comments

• [Datasets] Move `_add_partitions` logic to `TableBlockBuilder`

  #28868 commented on Feb 20, 2023 • 0 new comments

• [RLlib][OfflineRL] Cannot simultaneously have loader_fn for train data and have input_config with `format` for evaluation

  #29514 commented on Feb 20, 2023 • 0 new comments

• [RLlib] Add an informative error when gym version is not correct

  #29675 commented on Feb 20, 2023 • 0 new comments

• [core] prototype group-based oom worker killing policy

  #30014 commented on Feb 20, 2023 • 0 new comments

• [Datasets] Should remove actor-based prefetcher after #30375 is addressed

  #30380 commented on Feb 20, 2023 • 0 new comments

• [RLlib] rllib-cli examples fail: `stop` conditions are apparently no longer permitted in YAML?

  #31453 commented on Feb 20, 2023 • 0 new comments

• [Datasets] Calling limit() materializes data

  #29122 commented on Feb 20, 2023 • 0 new comments

• [air][data] unable to restore Datasets that call `map_batches`

  #29595 commented on Feb 20, 2023 • 0 new comments

• [air][data] unable to restore Datasets that call `repartition`

  #29597 commented on Feb 20, 2023 • 0 new comments

• [Datasets] Promote PandasRow and ArrowRow as PublicAPI

  #29761 commented on Feb 20, 2023 • 0 new comments

• [core] let raylet decide whether task / actor should be retried when it dies due to OOM

  #30013 commented on Feb 20, 2023 • 0 new comments

• [data] deprecate node:ip resources for data if any

  #31049 commented on Feb 20, 2023 • 0 new comments

• [rllib] remove node:ip resources and use NodeAffinitySchedulingPolicy

  #31052 commented on Feb 20, 2023 • 0 new comments

• Numpy 1.24 Support

  #31246 commented on Feb 20, 2023 • 0 new comments

• [RLlib] CQL Release Test is failing

  #31720 commented on Feb 20, 2023 • 0 new comments

• [data] [streaming] Large scale testing using MLPerf inference benchmark.

  #31796 commented on Feb 20, 2023 • 0 new comments

• [core] Getting node IP address by object ref

  #13630 commented on Feb 20, 2023 • 0 new comments

• [Serve Autoscaler] Add debug logs

  #20945 commented on Feb 20, 2023 • 0 new comments

• Update serve deployment FastAPI docs to make clear all steps needed to update endpoint

  #22719 commented on Feb 20, 2023 • 0 new comments

• [Documentation][autoscaler] Document the rest of autoscaler sdk, declare stable.

  #22815 commented on Feb 20, 2023 • 0 new comments

• passing a node as arg to method won't work

  #23057 commented on Feb 20, 2023 • 0 new comments

• [Datasets] [Testing] Add more nightly large-data-scale stress tests for Datasets

  #24391 commented on Feb 20, 2023 • 0 new comments

• [Serve] Don't pass deployment schema dict to DeploymentNode

  #24619 commented on Feb 20, 2023 • 0 new comments

• ray autoscaler not de-scaling when individual workers fixed workload completes

  #24844 commented on Feb 20, 2023 • 0 new comments

• [Core][Serve] Log statements missing in Ray v1.12 or newer

  #26039 commented on Feb 20, 2023 • 0 new comments

• [autoscaler] Better documentation on the node provider interface

  #26087 commented on Feb 20, 2023 • 0 new comments

• [Datasets] Revamp Datasets execution semantics and code

  #26264 commented on Feb 20, 2023 • 0 new comments

• [Datasets] `RobustScaler` doesn't handle divide-by-zero correctly

  #26857 commented on Feb 20, 2023 • 0 new comments

• [Core] [Documentation] Ray Up cloud infra assumptions and what we create

  #27326 commented on Feb 20, 2023 • 0 new comments

• [Datasets] Consolidate exception/error in Datasets

  #27914 commented on Feb 20, 2023 • 0 new comments

• Needs flags to enable/disable printing colored log messages

  #6855 commented on Feb 20, 2023 • 0 new comments

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

  #27863 commented on Feb 20, 2023 • 0 new comments

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

  #28071 commented on Feb 20, 2023 • 0 new comments

• [Core][utilization] some anti-pattern that not well supported by Ray core.

  #32297 commented on Feb 20, 2023 • 0 new comments

• [Python 3.11] add docs for python 3.11 wheels

  #31942 commented on Feb 20, 2023 • 0 new comments

• Class attribute in ray actor cannot be accessed/copied

  #11959 commented on Feb 20, 2023 • 0 new comments

• [core] occasional Ray port conflict issue

  #18053 commented on Feb 20, 2023 • 0 new comments

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

  #21419 commented on Feb 20, 2023 • 0 new comments

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

  #24020 commented on Feb 20, 2023 • 0 new comments

• [Core] race condition between job exits and actor creation.

  #24890 commented on Feb 20, 2023 • 0 new comments

• [core] Sending large binary blobs over object store scaler superlinearly

  #25217 commented on Feb 20, 2023 • 0 new comments

• [core] Scheduler stalls during shuffle reduce stage with 100k concurrent tasks or more

  #25412 commented on Feb 20, 2023 • 0 new comments

• [Ray clusters][Ray core][Docs][K8s] Document network communication of Ray components clearly.

  #25565 commented on Feb 20, 2023 • 0 new comments

• [Bug] `import ray` fails with `_Py_GetArgcArgv` symbol not found on macOS with Python 3.8.9

  #27380 commented on Feb 20, 2023 • 0 new comments

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

  #29127 commented on Feb 20, 2023 • 0 new comments

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

  #29466 commented on Feb 20, 2023 • 0 new comments

• [Core] Worker leak

  #30731 commented on Feb 20, 2023 • 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 Feb 20, 2023 • 0 new comments

• [Devprod] Bazel reports an error when compiling as a non-root user

  #30885 commented on Feb 20, 2023 • 0 new comments

• [core] Starting new thread in a task would result in a new task id

  #31879 commented on Feb 20, 2023 • 0 new comments

• [RFC][Serve] Multi Applications in serve 2.x API

  #31072 commented on Feb 20, 2023 • 0 new comments

• [core] `ray.get(future, timeout=0)` blocks forever but should return immediately

  #28465 commented on Feb 20, 2023 • 0 new comments

• Doing import ray breaks my logging [Bug]

  #22312 commented on Feb 20, 2023 • 0 new comments

• [Core] Async actor check failures when non-async method is received after exit_actor

  #32376 commented on Feb 20, 2023 • 0 new comments

• [core] potential issues when workers are not properly returned to worker pool

  #32271 commented on Feb 20, 2023 • 0 new comments

• [Core] ObjectStore fail to pull object, possibly because node info is missing

  #32046 commented on Feb 20, 2023 • 0 new comments

• [python3.11] Ray object refs can't be awaited using `asyncio.wait`

  #31606 commented on Feb 20, 2023 • 0 new comments

• [core] Objects created by num_returns="dynamic" temporarily leak if generator errors

  #28686 commented on Feb 20, 2023 • 0 new comments

• [Core] Loads report failed to catch killed driver.

  #25429 commented on Feb 20, 2023 • 0 new comments

• [Core/Clusters] Guidance on how to size the Ray head

  #25410 commented on Feb 20, 2023 • 0 new comments

• [Doc] Improve documentation on external Redis

  #24985 commented on Feb 20, 2023 • 0 new comments

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

  #25774 commented on Feb 20, 2023 • 0 new comments

• [Ray component: Core] Respect the placement group when `num_gpus=0, num_cpus=0`

  #27931 commented on Feb 20, 2023 • 0 new comments

• [Core] Add worker startup + module loading time to microbenchmark

  #30267 commented on Feb 20, 2023 • 0 new comments

• [core] Remove numpy from core dependencies

  #31249 commented on Feb 20, 2023 • 0 new comments

• [Core] Move `//python/ray/tests:test_tensorflow` from owned by core to owned by rllib

  #31250 commented on Feb 20, 2023 • 0 new comments

• [Core] Upgrade bazel to latest version (cur:4.2, latest:6.0)

  #31542 commented on Feb 20, 2023 • 0 new comments

• [Core] Application metrics use incorrect typing

  #32208 commented on Feb 20, 2023 • 0 new comments

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

  #28216 commented on Feb 20, 2023 • 0 new comments

• Don't set SIGTERM handler on import

  #17745 commented on Feb 20, 2023 • 0 new comments

• ray.init fails when temp_dir too long (but still reasonable)

  #7724 commented on Feb 20, 2023 • 0 new comments

• [Core][Feature request] Promote `actor_handle._actor_id` to public API

  #32638 commented on Feb 20, 2023 • 0 new comments

• [Job] Job supervisor logs are not printed to worker output

  #32377 commented on Feb 21, 2023 • 0 new comments

• Add warning when task backlog is too large

  #28051 commented on Feb 21, 2023 • 0 new comments

• [core] Scheduling a future on an actor with removed placement group hangs forever

  #28450 commented on Feb 21, 2023 • 0 new comments

• [data / core] Make it easier to debug system-level task failures for long-running jobs

  #29570 commented on Feb 21, 2023 • 0 new comments

• [core] replace node:ip resources by use NodeAffinitySchedulingPolicy

  #31051 commented on Feb 21, 2023 • 0 new comments

• [Core] [data] segfault with `ray.init(local_mode=True)` and `ray.data.range`

  #31160 commented on Feb 21, 2023 • 0 new comments

• [Core][docs] document nested task resource yielding behavior

  #2618 commented on Feb 21, 2023 • 0 new comments

• [core] How to obtain memory occupied by object ref?

  #14250 commented on Feb 21, 2023 • 0 new comments

• [Datasets] Truncate the plan properly in case it's too long for `Dataset.__repr__`

  #31744 commented on Feb 21, 2023 • 0 new comments

• [wheels] Remove unnecessary files from wheels

  #31751 commented on Feb 21, 2023 • 0 new comments

• [Core] `ray status` not consistent when using placement groups

  #25881 commented on Feb 21, 2023 • 0 new comments

• [AIR/docs] Extend Checkpoint section in AIR key concepts

  #31065 commented on Feb 21, 2023 • 0 new comments

• [tune/train] add some telemetry

  #32231 commented on Feb 21, 2023 • 0 new comments

• [Feature] [core] Selecting network interface

  #22732 commented on Feb 21, 2023 • 0 new comments

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

  #32067 commented on Feb 21, 2023 • 0 new comments

• [CI] `//python/ray/data:tests/test_object_gc` is flaky in master

  #32168 commented on Feb 21, 2023 • 0 new comments

• [Tune] Working with checkpoints user guide

  #32659 commented on Feb 22, 2023 • 0 new comments

• [core][state] ray log supporting regex searching

  #31549 commented on Feb 22, 2023 • 0 new comments

• [State Observability] Improve ray list job implementation.

  #26832 commented on Feb 22, 2023 • 0 new comments

• [Core][State Observability][Log] gRPC max size limit when `ray logs --tail 1000000`

  #27009 commented on Feb 22, 2023 • 0 new comments

• [Core] timeline doesn't show all infos.

  #28320 commented on Feb 22, 2023 • 0 new comments

• [Observability] No way to get head node ID

  #29607 commented on Feb 22, 2023 • 0 new comments

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

  #29959 commented on Feb 22, 2023 • 0 new comments

• [core][state] Track task events by job

  #31071 commented on Feb 22, 2023 • 0 new comments

• [core][state][log] ray log should traverse old directory for worker logs

  #31126 commented on Feb 22, 2023 • 0 new comments

• [core][state] Refactor use of bounded LRU/FIFO buffer/map used in task backend

  #31158 commented on Feb 22, 2023 • 0 new comments

• [core][state] Enforce max number of profiling events for one task.

  #31279 commented on Feb 22, 2023 • 0 new comments

• [core][state] Better report of data loss to users for task events

  #31280 commented on Feb 22, 2023 • 0 new comments

• core - ray logs CLI doesn't work for kubernetes raycluster

  #31381 commented on Feb 22, 2023 • 0 new comments

• [core][doc] Expired link to slack invitation in state API

  #31536 commented on Feb 22, 2023 • 0 new comments

• [core][state] Microbench regression with task backend investigation

  #31546 commented on Feb 22, 2023 • 0 new comments

• [core][state] Surface task log for each task from state apis

  #31611 commented on Feb 22, 2023 • 0 new comments

• [core][state] Actor creation task marked as FINISHED rather than FAILED

  #31620 commented on Feb 22, 2023 • 0 new comments

• [core][state] Human readable display of state API output

  #31876 commented on Feb 22, 2023 • 0 new comments

• [core][job] Change job related timestamp to use nanoseconds

  #32163 commented on Feb 22, 2023 • 0 new comments

• [State API] ray log truncation message improvements

  #32392 commented on Feb 22, 2023 • 0 new comments

• [Dataset] Add logical operator for `zip()`

  #32597 commented on Feb 22, 2023 • 0 new comments

• [Datasets] Enabling execution optimizer after passing all CI and nightly tests

  #32596 commented on Feb 22, 2023 • 0 new comments

• [Core][State Observability] Grouping state APIs when `ray --help` is called.

  #26376 commented on Feb 22, 2023 • 0 new comments

• [Core] Worker exit is not reported if the worker is dead by node exit

  #24957 commented on Feb 22, 2023 • 0 new comments

• [Dataset] Pass initial stats of plan to executor

  #32601 commented on Feb 22, 2023 • 0 new comments

• [Datasets] Validate aggregate function in logical operator.

  #32463 commented on Feb 22, 2023 • 0 new comments

• [Datasets] Validate sort key in logical operator

  #32137 commented on Feb 22, 2023 • 0 new comments

• [Datasets] Support non-shuffle repartition in logical operator

  #32108 commented on Feb 22, 2023 • 0 new comments

• [Datasets] Run `Dataset.schema()` with logical operator and executor

  #32667 commented on Feb 22, 2023 • 0 new comments

• [Datasets] Allow fuse MapOperator with AllToAllOperator

  #32608 commented on Feb 22, 2023 • 0 new comments

• [Core] Importing ray modifies `sys.path`, resulting in unexpected behaviour

  #32183 commented on Feb 22, 2023 • 0 new comments

• [Core] Process signalling using `SIGINT` not working in `ray.remote` processes

  #31805 commented on Feb 22, 2023 • 0 new comments

• [core] Improve serialization error for tasks, actors, and put

  #31074 commented on Feb 22, 2023 • 0 new comments

• [Core] Ray doesn't shutdown properly on KeyboardInterrupt

  #29384 commented on Feb 22, 2023 • 0 new comments

• [Datasets] Add a new `"zero_copy"` batch format

  #32662 commented on Feb 23, 2023 • 0 new comments

• [data] [streaming] Enable streaming executor by default

  #32132 commented on Feb 23, 2023 • 0 new comments

• [Core] Long-running workload produces >200K log files on a single node in `/tmp/ray/session_latest/logs`

  #32112 commented on Feb 23, 2023 • 0 new comments

• [Core] Certain log files cannot be followed/streamed

  #29928 commented on Feb 24, 2023 • 0 new comments

• [core] oom killer: kill task more aggressively

  #29664 commented on Feb 24, 2023 • 0 new comments

• [cluster][doc] Add AWS instructions for GPU training tutorial

  #28901 commented on Feb 24, 2023 • 0 new comments

• [Serve] k8s release tests show success when test fails.

  #32668 commented on Feb 24, 2023 • 0 new comments

• [Serve] annotation for tag_keys is wrong

  #32494 commented on Feb 24, 2023 • 0 new comments

• Runtime Environment Dependencies- container per task

  #28875 commented on Feb 24, 2023 • 0 new comments

• [<Algorithm overview>]

  #28915 commented on Feb 24, 2023 • 0 new comments

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

  #29068 commented on Feb 24, 2023 • 0 new comments

• [Core] some workloads were not able to be distributed on cluster.

  #30200 commented on Feb 24, 2023 • 0 new comments

• [Ray release infra] Flaky tests caused by autoscaling group killing instances while tests are running

  #32013 commented on Feb 24, 2023 • 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 Feb 24, 2023 • 0 new comments

• [Core] Remove actor deletion upon job termination

  #31019 commented on Feb 24, 2023 • 0 new comments

• [core] Fix demand leak when worker failed

  #31175 commented on Feb 19, 2023 • 0 new comments

• [After 2.3 branch cut] [Core] Make multiprocessing pool raise TypeError when provided a non-iterable in imap or imap_unordered

  #31799 commented on Feb 22, 2023 • 0 new comments

• [docs] algolia v3 docsearch

  #31992 commented on Feb 21, 2023 • 0 new comments

• [Datasets] [Operator Fusion - E2E Mono-PR] [DO-NOT-MERGE] Add zero-copy operator fusion.

  #32178 commented on Feb 22, 2023 • 0 new comments

• [build_base] [don't review] fix

  #32304 commented on Feb 25, 2023 • 0 new comments

• [Core][Enable gcs scheduler 8/n] Releasing gcs actor scheduler

  #32314 commented on Feb 23, 2023 • 0 new comments

• [WIP - dont' review] Enable streaming executor by default

  #32493 commented on Feb 25, 2023 • 0 new comments

• [RLLib] CNN and Catalog

  #32650 commented on Feb 23, 2023 • 0 new comments

• [WIP] [data] [streaming] Introduce output split operator

  #32675 commented on Feb 22, 2023 • 0 new comments

• [core][oom] Lower oom threshold

  #32676 commented on Feb 23, 2023 • 0 new comments