Commit Graph

2901 Commits

Author SHA1 Message Date
Kai Fricke 6b6780a108 [tune] Allow trials to remain paused in BOHB (#10531)
Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
2020-09-04 15:29:33 -07:00
Eric Liang da83bbd764 [1.0] Move dask scheduler from experimental to util (#10553)
* move dask

* fix dask
2020-09-04 12:16:32 -07:00
Edward Oakes b1bd58581b Print "Connecting to existing cluster" message to stderr instead of stdout (#10575) 2020-09-04 12:51:59 -05:00
Maksim Smolin 02efa2549d [cli] Fix colors on light themes (#10577) 2020-09-04 10:45:27 -07:00
Edward Oakes 786f12edfd [serve] Serve client refactor (#10409) 2020-09-04 12:02:23 -05:00
Kai Fricke 2e49e22f21 [tune] Add test_sample to bazel BUILD (#10566) 2020-09-04 09:09:02 -07:00
Kai Yang 5f5160ead9 [Core] Multi-tenancy: Worker capping (#10500) 2020-09-04 20:34:06 +08:00
PidgeyBE 51df0820de [autoscaler] Fix ingress manifest bug (#10536) 2020-09-04 01:09:16 -05:00
Kai Fricke 5c3d4a6670 [tune] added MXNet integration callbacks (#10533) 2020-09-03 18:06:44 -07:00
Edward Oakes ead30ca655 [Core] fix named actor bug (#10550) 2020-09-03 17:48:31 -07:00
Simon Mo 94374e1dd9 [Serve] Add Latency and Queue Size Metrics (#10535) 2020-09-03 17:33:37 -07:00
Simon Mo eff4375c3d [Serve] Produtionize Starlette Middlewares (#10529) 2020-09-03 17:31:38 -07:00
architkulkarni 0d93e92720 [Serve] Reimplement BackendConfig as pydantic model (#10389) 2020-09-03 19:16:17 -05:00
Richard Liaw 43a7a64b30 [tune] horovod trainable (#10304) 2020-09-03 16:53:35 -07:00
Clark Zinzow 7068c63dd8 Set Ray task name to Dask key for Dask tasks. (#10547) 2020-09-03 15:37:55 -07:00
Ian Rodney c54853d45b [Autoscaler] Actually try to catch when docker does not exist (#10549) 2020-09-03 14:00:06 -07:00
Ian Rodney a13c83d7f0 Add WorkerCrashedError to cancel docs (#10534) 2020-09-03 13:23:04 -07:00
Clark Zinzow 0c0b0d0a73 [Core] Added support for submission-time task names. (#10449)
* Added support for submission-time task names.

* Suggestions from code review: add missing consts

Co-authored-by: SangBin Cho <rkooo567@gmail.com>

* Add num_returns arg to actor method options docstring example.

* Add process name line and proctitle assertion to submission-time task name section of advanced docs.

* Add submission-time task name --> proctitle test for Python worker.

* Added Python actor options tests for num_returns and name.

* Added Java test for submission-time task names.

* Add dashboard image to task name docs section.

* Move to fstrings.

Co-authored-by: SangBin Cho <rkooo567@gmail.com>
2020-09-03 11:45:24 -07:00
Edward Oakes 71274954d1 Remove unnecessary output when connecting to a cluster. (#10512) 2020-09-03 13:30:33 -05:00
Edward Oakes e4d80e1b0f fix passing sys config to start (#10514) 2020-09-03 11:18:21 -07:00
krfricke 91535e9102 [tune] Refactored Keras integration callbacks (#10509) 2020-09-03 10:16:08 -07:00
krfricke 06af62ba91 [tune] refactor tune search space (#10444)
* Added basic functionality and tests

* Feature parity with old tune search space config

* Convert Optuna search spaces

* Introduced quantized values

* Updated Optuna resolving

* Added HyperOpt search space conversion

* Convert search spaces to AxSearch

* Convert search spaces to BayesOpt

* Added basic functionality and tests

* Feature parity with old tune search space config

* Convert Optuna search spaces

* Introduced quantized values

* Updated Optuna resolving

* Added HyperOpt search space conversion

* Convert search spaces to AxSearch

* Convert search spaces to BayesOpt

* Re-factored samplers into domain classes

* Re-added base classes

* Re-factored into list comprehensions

* Added `from_config` classmethod for config conversion

* Applied suggestions from code review

* Removed truncated normal distribution

* Set search properties in tune.run

* Added test for tune.run search properties

* Move sampler initializers to base classes

* Add tune API sampling test, fixed includes, fixed resampling bug

* Add to API docs

* Fix docs

* Update metric and mode only when set. Set default metric and mode to experiment analysis object.

* Fix experiment analysis tests

* Raise error when delimiter is used in the config keys

* Added randint/qrandint to API docs, added additional check in tune.run

* Fix tests

* Fix linting error

* Applied suggestions from code review. Re-aded tune.function for the time being

* Fix sampling tests

* Fix experiment analysis tests

* Fix tests and linting error

* Removed unnecessary default_config attribute from OptunaSearch

* Revert to set AxSearch default metric

* fix-min-max

* fix

* nits

* Added function check, enhanced loguniform error message

* fix-print

* fix

* fix

* Raise if unresolved values are in config and search space is already set

Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
2020-09-03 09:06:13 -07:00
chaokunyang ea95e6f7cc [Java] lint java code (#10494) 2020-09-03 10:39:14 +08:00
Ian Rodney b9633a2b67 [docker] Support multiple node types (#10504) 2020-09-02 18:27:59 -07:00
SangBin Cho dc7fe1a4c5 [Placement Group] Atomic Placement Group Part 1, Basic Structure. (#10482)
* Write a test.

* Basic structure done.

* Reduce flakiness of tests.

* Addressed code review.

* Skipping tests because it is flaky for now.

* Fix linting issues.

* Increase sleep time to see lint messages.

* Lint issue fixed.
2020-09-02 18:14:46 -07:00
Ian Rodney 4324dd5929 [docker] Refactor "autoscaler" image into "-autoscaler" tag and "ray-ml" image. (#10351) 2020-09-02 13:03:35 -07:00
krfricke 57c4183724 [tune] add xgboost callbacks to integration module (#10502) 2020-09-02 11:16:09 -07:00
Vysybyl 6fa0edfbef [gcp] Update config.py for safe dir creation (#9645)
Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
2020-09-01 21:41:44 -07:00
fyrestone b04222dbd9 [xlang] Cross language serialization for ActorHandle (#10335) 2020-09-02 10:11:53 +08:00
Simon Mo 65f17f2e14 [Serve] Refactor RequestMetadata and Query objects (#10483) 2020-09-01 18:15:31 -07:00
raoul-khour-ts 3b10b67a15 [tune] SigOpt multi-objective search + experiments (#10457) 2020-09-01 16:22:29 -07:00
Yiran Wang 2b95b613f2 [Autoscaler] Retry create_instances properly in AWSNodeProvider (#10479) 2020-09-01 16:17:11 -07:00
Alex Wu 23bbe0f36a [Autoscaler] Reload config (#10450) 2020-09-01 14:37:04 -07:00
krfricke 1dd55f4b07 [tune] remove callbacks from config in wandb logger initialization (#10441) 2020-09-01 14:26:39 -07:00
Richard Liaw 3f98a8bfcb [docs] Fix warnings for sphinx 1.8 (#10476)
* fix-build-for-sphinx18

* jnilit
2020-09-01 13:37:35 -07:00
Ian Rodney 283f4d1060 [docker] Use tmp paths for rsync and fix file_mounts on docker (#10368) 2020-09-01 13:14:35 -07:00
Ian Rodney c644650818 [docker] Run docker stop in teardown_cluster (#10407) 2020-09-01 11:25:37 -07:00
Richard Liaw 09d4a3241f [tune] Support true pooling and batched concurrency (#10352) 2020-09-01 10:33:49 -07:00
Simon Mo ddd62a177f Revert "[Serialization] Update CloudPickle to 1.6.0 (#9694)" (#10460)
This reverts commit f0c3910d59.
2020-08-31 20:41:37 -07:00
krfricke f3f698816d [tune] Added PyTorch Lightning callbacks to integrations (#10220)
Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
2020-08-31 15:30:48 -07:00
Richard Liaw d8e7b144e4 [tune] avoid distributed error message (#10387) 2020-08-31 15:29:43 -07:00
Jan Blumenkamp 5774e1ce88 [tune/rllib/wandb] Flatten result dict so that nested result dicts are shown in W&B logger (#10429) 2020-08-31 15:28:46 -07:00
Sumanth Ratna bac8f14739 [tune] Use isinstance instead of type in assert in HyperOptSearch (#10454) 2020-08-31 15:26:55 -07:00
krfricke efa1d51aea [tune] Added external PyTorch tutorial test (#10192) 2020-08-31 15:24:31 -07:00
Zhe Zhang 4356abeb1c [Doc] Fix errors in ActorPool documentation (#10410)
* Fix errors in ActorPool documentation

1. map() and map_unordered() name
2. The print statement doesn't work

* Correctly change lines

* Address comments from pr
2020-08-31 13:57:10 -07:00
raoul-khour-ts 25f5614691 [tune] Rrk/sigopt_searcher_improvements (#10446) 2020-08-31 13:15:12 -07:00
PidgeyBE 6917efabc4 [k8s] Replace k8s service name everywhere in ingress manifest (#10445)
Co-authored-by: Pieterjan <pieterjan.soetaert@robovision.eu>
2020-08-31 13:14:40 -07:00
Robert Nishihara 0bba5485d3 [cli] Add prompt command for CLI logger. (#9897)
Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
2020-08-31 11:03:53 -07:00
Amog Kamsetty afde3db4f0 [Tune] Synchronous Mode for PBT (#10283) 2020-08-31 00:00:47 -07:00
SangBin Cho 3e5cac80d8 [Tests] Fix Broken GCS restart test. (#10417) 2020-08-30 00:44:34 -07:00