Commit Graph

152 Commits

Author SHA1 Message Date
Amog Kamsetty f87a4aa45d [Tune] Pbt Function API (#9958)
* adding function convnet example

* add unit test

* update test

* update example

* wip

* move error from experiment to tune

* wip

* Fix checkpoint deletion

* updating code

* adding smoke test

* updating pbt guide

* formatting

* fix build

* add best checkpoint analysis util

* update test

* add comments

* remove class api

* fix example

* add setup and teardown to tests

* formatting

* Update python/ray/tune/tests/test_trial_scheduler_pbt.py

Co-authored-by: Kai Fricke <kai@anyscale.com>
Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
2020-08-14 17:52:30 -07:00
krfricke 16486a8df3 [tune] Add OptunaSearcher wrapper around Optuna samplers (#10044)
Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
Co-authored-by: Kai Fricke <kai@anyscale.com>
2020-08-12 16:13:22 -07:00
krfricke 221fdc0774 [tune] fix flaky PBT replay test (#10047)
Co-authored-by: Kai Fricke <kai@anyscale.com>
2020-08-11 14:17:31 -07:00
Richard Liaw 98df612010 [tune] option to raise on error (#10030) 2020-08-11 09:59:04 -07:00
Amog Kamsetty 856d4a0533 [Tune] Better error when using checkpoint_freq (#9998) 2020-08-10 13:52:46 -07:00
Richard Liaw be8e63d477 [tune] support resume for search algorithms (#9972) 2020-08-10 13:43:14 -07:00
krfricke 0ef8224446 [tune] PBT replay utility scheduler (#9953)
Co-authored-by: Kai Fricke <kai@anyscale.com>
2020-08-07 12:41:49 -07:00
Richard Liaw c6404e8cf6 [tune] Search alg checkpointing during training (#9803)
Co-authored-by: krfricke <krfricke@users.noreply.github.com>
2020-08-03 15:07:31 -07:00
Richard Liaw b5068d08bf [tune] Fix restoration for function API PBT (#9853) 2020-08-03 12:36:17 -07:00
krfricke c741d1cf9c [tune] stdout/stderr logging redirection (#9817)
* Add `log_to_file` parameter, pass to Trainable config, redirect stdout/stderr.

* Add logging handler to root ray logger

* Added test for `log_to_file` parameter

* Added logs, reuse test

* Revert debug change

* Update logdir on reset, flush streams after each train() step

* Remove magic keys from visible config

Co-authored-by: Kai Fricke <kai@anyscale.com>
2020-08-03 11:18:34 -07:00
Richard Liaw a47121476f [tune] Remove accidentally added files (#9835) 2020-07-30 21:47:27 -07:00
krfricke 619e44e54a [tune] Added WandbLogger (#9725)
Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
Co-authored-by: Kai Fricke <kai@anyscale.com>
2020-07-30 13:09:03 -07:00
Richard Liaw 0c3b9ebeef [tune/sgd] Document func_trainable and add checkpoint context (#9739)
Co-authored-by: krfricke <krfricke@users.noreply.github.com>
Co-authored-by: Amog Kamsetty <amogkam@users.noreply.github.com>
2020-07-30 09:46:37 -07:00
krfricke ad0219b80d [tune] fix pbt checkpoint_freq (#9517)
* Only delete old checkpoint if it is not the same as the new one

* Return early if old checkpoint value coincides with new checkpoint value

Co-authored-by: Kai Fricke <kai@anyscale.com>
2020-07-18 00:58:16 -07:00
krfricke 87630cf024 [tune] Unflattened lookup for ProgressReporter (#9525)
Co-authored-by: Kai Fricke <kai@anyscale.com>
2020-07-17 13:52:54 -07:00
Michael Mui e93cde8c66 [tune] Issue 8821: ExperimentAnalysis doesn't expand user (#9461) 2020-07-14 13:53:37 -07:00
Nicolaus93 b5a6c57295 [tune] handling nan values (#9381) 2020-07-12 17:08:36 -07:00
Richard Liaw b71c912da7 [tune] Fix up examples (#9201) 2020-07-05 01:16:20 -07:00
David Fidalgo c0ba337fe0 [tune] Add np.bool8 and np.int to allowed HPARAMS types (#9297) 2020-07-03 18:34:45 -07:00
Richard Liaw d35f0e40d0 [tune] Use public methods for trainable (#9184) 2020-07-01 11:00:00 -07:00
mehrdadn 898e472425 Make test_utils.py use pipes to avoid file access conflicts on Windows (#9072)
Co-authored-by: Mehrdad <noreply@github.com>
2020-06-27 22:33:45 +02:00
Richard Liaw e6ee39a6a3 [tune] checkpoint_dir test (#8024) 2020-06-20 17:56:24 -07:00
mehrdadn 981f67bfb0 Fix more Windows issues (#9011)
Co-authored-by: Mehrdad <noreply@github.com>
2020-06-19 18:51:45 -07:00
Richard Liaw 6c49c01837 [tune] Function API checkpointing (#8471)
Co-authored-by: krfricke <krfricke@users.noreply.github.com>
2020-06-15 10:42:54 -07:00
krfricke 060e524c92 [tune] Parameter columns can now be specified in tune reporters (#8802)
Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
Co-authored-by: Kai Fricke <kai@anyscale.com>
2020-06-11 11:30:25 -07:00
Richard Liaw fc54dc8652 [tune] Make test_api faster (#8844) 2020-06-09 12:45:27 -07:00
Richard Liaw d7b64ef279 [tune] BayesOpt - finish early when optimizer converges (#8808) 2020-06-09 12:09:39 -07:00
Amog Kamsetty 9410e5884d [Tune] Parametrize Cloud Syncing Frequency (#8771) 2020-06-04 18:55:50 -07:00
krfricke f4ee3e76d8 [tune] last-n-avg
Co-authored-by: Kai Fricke <kai@anyscale.com>
2020-06-02 20:06:04 -07:00
Sven Mika d483ed28ba [RLlib] Fix broken tune tests in master due to framework=auto errors. (#8672) 2020-05-29 11:55:47 +02:00
Luca Cappelletti c9898eff24 [Tune] Added method to integrate previous analysis in BO (#8486) 2020-05-19 23:26:43 -07:00
Luca Cappelletti 5b330de182 [Tune] Introduced patience to early stopping (#8484) 2020-05-18 13:12:16 -07:00
Luca Cappelletti d1ef70da16 [Tune] Added default values for utility kwargs (#8488) 2020-05-18 13:10:43 -07:00
Luca Cappelletti 2ff26f13d2 [tune] Added EarlyStopping and relative test suite (#8459) 2020-05-17 12:18:59 -07:00
Richard Liaw 67c01455fe [tune] tune.track -> tune.report (#8388) 2020-05-16 12:55:08 -07:00
krfricke 4633d81c39 [tune] added average scope to experiment analysis (#8445) 2020-05-14 15:20:43 -07:00
Eric Liang 1126fe4d23 [tune] Add UUID back to trial names (#8377) 2020-05-08 20:20:36 -07:00
Richard Liaw 40dfb337bf [tune] Hotfix Ax breakage when fixing backwards-compat (#8285) 2020-05-02 20:42:50 -07:00
Richard Liaw 4d639354cd [tune] Hotfix for test_ls (#8215) 2020-04-28 14:06:12 -07:00
Richard Liaw 87557a00fa [tune] Refactor search algorithms (#7037)
* start refactoring of search algorithms

* format

* needs tests

* fix

* suggestions

* Fix PBT

* lint

* refactoring

* hyperopt_working

* dragonfly

* hyperopt

* change_half_of_algs

* save

* code-removed

* remove_lots_of_unneccessary

* changes

* formatting

* suggest

* reset

* rm

* tests

* search-change

* exception

* refactor-doc

* search

* py

* moredocs

* Update doc/source/tune-searchalg.rst

* concurrency

* max

* tune

* betterwarning

* bohb

* tests

* test-change

Co-authored-by: ujvl <misraujval@gmail.com>
2020-04-27 08:51:13 -07:00
Richard Liaw 9dd3490c38 [tune] Safer try-catch for TensorboardX (#8174)
Co-Authored-By: Kristian Hartikainen <kristian.hartikainen@gmail.com>
2020-04-25 13:08:37 -07:00
Ujval Misra 708dff6d8f [tune] Stop-gap fix for PBT checkpointing (#7794)
* Fix PBT

* lint

* reset

* rm

* tests

Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
2020-04-20 15:10:36 -07:00
Richard Liaw de1787e5e5 [tune] Check actor start -> test_cluster (#8056)
* test

* info

* ok

* hard_stop

* codefix
2020-04-16 20:00:45 -07:00
Richard Liaw 4d8bf5635d [hotfix] Lint formatting for new Tune optimizer ZOOpt (#8040)
* formatting

* removedill

* lint
2020-04-16 09:24:30 -07:00
Servon 5c274fe631 [Tune] Add ZOOpt search algorithm (#7960)
* add zoopt

* add zoopt search algo

* add zoopt

* fix zoopt

* add zoopt requirements

* fix zoopt

* remove generated guides

* Apply suggestions from code review

Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
2020-04-15 21:13:29 -07:00
Richard Liaw 87e3c39b48 [tune] Ensure Cleanup (#7967) 2020-04-11 16:28:03 -07:00
Richard Liaw ca6eabc9cb [tune] Fail Fast (#7528)
* pytest

* init cancel

* testing

* Update python/ray/tune/tests/test_tune_server.py

Co-Authored-By: Richard Liaw <rliaw@berkeley.edu>

* change-test

* Apply suggestions from code review

* Apply suggestions from code review

* finished

* set_finished

* tune

* fix

Co-authored-by: ijrsvt <ian.rodney@gmail.com>
2020-03-26 00:04:09 -07:00
Richard Liaw 82b792be33 [tune] IP Check, Flatten Results for TBX (#7705)
* support_flattened

* loggers

* Format logger changes

Co-authored-by: Kristian Hartikainen <kristian.hartikainen@gmail.com>
2020-03-25 09:18:03 +00:00
Richard Liaw 54a892bb84 [tune] Cancel Experiment via Client (#7719)
* init cancel

* testing

* Update python/ray/tune/tests/test_tune_server.py

Co-Authored-By: Richard Liaw <rliaw@berkeley.edu>

* Apply suggestions from code review

* Apply suggestions from code review

* finished

* set_finished

Co-authored-by: ijrsvt <ian.rodney@gmail.com>
2020-03-24 20:30:12 -07:00
Eric Liang 86f89fc3b3 [tune] Higher timeout for progress reporter test (#7679)
* wip

* medium size
2020-03-22 13:47:08 -07:00