Eric Liang
b1660c4edf
[rllib] Refactor to support passing custom env_creator function ( #1096 )
...
* refactor to use env creator
* doc
* lint
2017-10-10 12:49:42 -07:00
Robert Nishihara
8f1a73f041
Allow Ray to be built without UI by setting INCLUDE_UI=0. ( #1094 )
...
* Allow building Ray without UI by setting INCLUDE_UI=0.
* Fix bash.
* Fix linting.
2017-10-09 23:32:38 -07:00
Robert Nishihara
a52a1e893f
Automatically set CUDA_VISIBLE_DEVICES when worker gets task. ( #1044 )
...
* Automatically set CUDA_VISIBLE_DEVICES when worker gets task.
* Add test.
2017-10-06 18:38:08 -07:00
Robert Nishihara
4669c59fa8
Release GPU resources as soon as an actor exits. ( #1088 )
...
* Release GPU resources as soon as an actor exits.
* Add a test.
* Store local_scheduler_id and driver_id in the worker object instead of the actor object.
2017-10-06 17:58:19 -07:00
Stephanie Wang
aebe9f9374
Fix actor garbage collection by breaking cyclic references ( #1064 )
...
* Fix bug in wait_for_pid_to_exit, add test for actor deletion.
* Fix actor garbage collection by breaking cyclic references
* Add test for calling actor method immediately after actor creation.
* Fix bug, must dispatch tasks when workers are killed.
* Fix python test
* Fix cyclic reference problem by creating ActorMethod objects on the fly.
* Try simply increasing the time allowed for many_drivers_test.py.
2017-10-05 00:55:33 -07:00
Robert Nishihara
971becc905
[rllib] Use NoFilter instead of MeanStdFilter for PPO. ( #1082 )
...
* Make NoFilter the default observation filter for PPO.
* Make reward filter NoFilter for PPO.
2017-10-04 21:31:17 -07:00
Eric Liang
6ecc899cf2
[rllib] Fix DQN checkpoint/restore and enable test in jenkins ( #1063 )
...
* fix dqn restore and add test
* Update .gitignore
* Update test_checkpoint_restore.py
* add checkpoint restore
2017-10-03 23:17:54 -07:00
Richard Liaw
cb6dea94bc
[rllib] Fix Preprocessor for ATARI ( #1066 )
...
* Removing squeeze, fix atari preprocessing
* nit comment
* comments
* jenkins
* Lint
2017-10-03 18:45:02 -07:00
Philipp Moritz
57bd1d6ff5
Specialize Serialization for OrderedDict ( #1035 )
...
Specialize Serialization for OrderedDict and defaultdict
2017-10-02 17:33:10 -07:00
Philipp Moritz
b94d85fb5d
[rllib] Fix logging to Athena ( #1058 )
...
* Fix logging to Athena
* fixes
2017-10-02 17:16:52 -07:00
Richard Liaw
54aaeb25ab
fixing ( #1048 )
2017-10-01 22:34:55 -07:00
Robert Nishihara
2892955119
Pass string into json.loads, not bytes object. ( #991 )
2017-10-01 14:56:39 -07:00
Robert Nishihara
6d393828db
Bump version number to 0.2.1. ( #1026 )
2017-10-01 12:33:13 -07:00
Robert Nishihara
ad61af7333
Workaround for passing empty list to ray.wait. ( #1043 )
...
* Workaround for passing empty list to ray.wait.
* Add test for passing empty list to wait.
2017-10-01 11:45:02 -07:00
Augusto Yao
0d90a17426
Pass cleanup argument to start_monitor. ( #1040 )
2017-09-30 15:35:25 -07:00
Wapaul1
97b3355adc
Register Class Only Creates Entry in Redis Once ( #1038 )
...
Don't export the same custom class definition multiple times.
2017-09-30 15:30:27 -07:00
Richard Liaw
16e82b43d1
[rllib] Changes for preprocessors ( #1033 )
...
* Changes for preprocessors
* removed comments
* Changes + push for lint
* linted
* adding dependency for travis
* linting won't pass
* reordering
* needed for testing
* added comments
* pip it
* pip dependencies
2017-09-30 13:11:20 -07:00
Alexey Tumanov
2d0f439b7b
hugepage + plasma directory support plumbing + documentation ( #1030 )
...
* hugepage + plasma directory support plumbing + documentation
* Indentation fix.
* huge_pages_enabled --> huge_pages
* One more change
2017-09-30 09:56:52 -07:00
Robert Nishihara
b991dc8900
Add flag for ignoring the UI, don't start UI in jenkins tests. ( #1021 )
2017-09-29 15:22:51 -07:00
Eric Liang
9f3a4fce50
[rllib] Parallelize sample collection and gradient computation in DQN ( #746 )
...
* wip
* works with cartpole
* lint
* fix pg
* comment
* action dist rename
* preprocessor
* fix test
* typo
* fix the action[0] nonsense
* revert
* satisfy the lint
* wip
* wip
* works with cartpole
* lint
* fix pg
* comment
* action dist rename
* preprocessor
* fix test
* typo
* fix the action[0] nonsense
* revert
* satisfy the lint
* Minor indentation changes.
* fix merge
* add humanoid
* initial dqn refactor
* remove tfutil
* fix calls
* fix tf errors 1
* closer
* runs now
* lint
* tensorboard graph
* fix linting
* more 4 space
* fix
* fix linT
* more lint
* oops
* es parity
* remove example.py
* fix training bug
* add cartpole demo
* try fixing cartpole
* allow model options, configure cartpole
* debug
* simplify
* no dueling
* avoid out of file handles
* Test dqn in jenkins.
* Minor formatting.
* lint
* fix py3
* fix issue
* remove chekcpoint
* revert
* Fixit
* sanity check configs
* update cuda
* fix
* parallel gradient computation
* update
* upd
* bug
* upd
* always record training stats
* fix
* comments
* revert assert
* add gpu mask
* fofset
* a tie
* Merge
* fix
* fix
* fix examples
* A3C -> DQN
* fix dqn test
* remove submodule
* fix linting
2017-09-29 00:06:51 -07:00
Eric Liang
19562f6ce5
[rllib] Fix issues with PPO model restoration ( #1018 )
...
* fix filter
* add test
* lint
* fix
* commit
* Update a3c.py
2017-09-28 13:12:06 -07:00
Zongheng Yang
5a50e80b63
Make Monitor remove dead Redis entries from exiting drivers. ( #994 )
...
* WIP: removing OL, OI, TT on client exit; no saving yet.
* ray_redis_module.cc: update header comment.
* Cleanup: just the removal.
* Reformat via yapf: use pep8 style instead of google.
* Checkpoint addressing comments (partially)
* Add 'b' marker before strings (py3 compat)
* Add MonitorTest.
* Use `isort` to sort imports.
* Remove some loggings
* Fix flake8 noqa marker runtest.py
* Try to separate tests out to monitor_test.py
* Rework cleanup algorithm: correct logic
* Extend tests to cover multi-shard cases
* Add some small comments and formatting changes.
2017-09-26 00:11:38 -07:00
Eric Liang
5c70faf76b
Update common.py ( #996 )
2017-09-19 10:10:56 -07:00
gycn
a432285e77
Disable parallelization for Actors and ray.wait for debugging ( #961 )
...
Support actors and ray.wait in PYTHON_MODE.
2017-09-17 00:12:50 -07:00
Philipp Moritz
73f40bd844
[rllib] user defined preprocessor ( #985 )
...
* add register_preprocessor to ModelCatalog
* add pytest
* make staticmethod a classmethod
* update
* install gym on travis
* fix linting
* fix
2017-09-16 15:53:19 -07:00
Eric Liang
98142ef51f
fix checkpoint ( #988 )
2017-09-16 15:29:36 -07:00
Philipp Moritz
6601bb5f9e
[rllib] Make observation filter optional ( #940 )
...
* make observation filter optional
* fix linting
2017-09-14 17:37:19 -07:00
Richard Liaw
d516d9440e
Fixing local directory ( #977 )
...
* Fixing local directory
Enables ability to set custom local directory; code may be messy.
* Create all intermediate parent directories
2017-09-14 10:33:52 -07:00
Philipp Moritz
1eb8c83314
[rllib] Initial RLLib documentation ( #969 )
...
* initial documentation for RLLib
* more RL documentation
* fix linting
* fix comments
* update
* fix
2017-09-12 23:38:21 -07:00
Eric Liang
9f42ef6a4f
[rllib] Make sure to always record stats like time elapsed, timesteps ( #965 )
...
* always record training stats
* fix
* comments
* revert assert
* nan
* fix
2017-09-12 14:28:16 -07:00
Eric Liang
e17412a72b
fix free log std param ( #964 )
2017-09-11 18:52:48 -07:00
Stephanie Wang
99c8b1f38c
Actor fault tolerance using object lineage reconstruction ( #902 )
...
* Revert Python actor reconstruction
* Actor reconstruction using object lineage
* Add dummy arguments and return values for actor tasks
* Pin dummy outputs for actor tasks
* Skip checkpointing test for now
* TODOs
* minor edits
* Generate dummy object dependencies in Python, not C
* Fix linting.
* Move actor counter and dummy objects inside of the actor handle
* Refactor Worker._process_task, suppress exception propagation for
sequential actor tasks
2017-09-10 19:29:28 -07:00
Eric Liang
d8aa826e63
[webui] Scalability fixes for the task timeline and visualizations ( #935 )
...
* fixes
* comments
* fix test
* Update ui.py
* upd
* Fix linting.
2017-09-10 15:47:44 -07:00
Robert Nishihara
f3c1248d98
Clone catapult and generate html files during installation. ( #956 )
...
* Clone catapult and generate static html during setup.
* Include UI files in installation.
* Fix directory to clone catapult to and fix linting.
* Use absolute path.
* Make sure we find a sufficiently new version of python2 when building wheels.
* Copy the trace_viewer_full.html file to the local directory if it is not present.
* Make sure wheels fail to build if UI is not included.
2017-09-10 13:41:16 -07:00
Philipp Moritz
546ba23ceb
Upgrade to latest arrow to include set serialization speedups ( #957 )
...
* update arrow to pull in the set serialization speedups
* remove _register_class for set
2017-09-10 00:12:17 -07:00
Eric Liang
953878364e
[webui] Print out timeline link for full-screen trace viewing ( #936 )
...
* up
* update
2017-09-06 01:41:21 -07:00
Wapaul1
e19e2c6284
Print jupyter notebook token when starting web UI. ( #887 )
...
* User now only needs to copy url to get to notebook
* Fixed duplicate code
* Added function to print url
* Added exception for calling function on worker
* Stored webui url in Redis
* Fix linting and simplify code.
* Now uses 24 bytes hex token
* Fixed python 3 compatibility
* Fix linting and python 3 compat
* Added comment explaining generating the token.
* Removed newline
* Small fixes.
* Fixed jenkins failure
* Rebased and changed formatting
* Revert "changed formatting"
This reverts commit 226510cf0cdcaab9cf42ad30bd9588a963683592.
2017-09-05 23:31:44 -07:00
Robert Nishihara
853969225b
Sleep longer when starting plasma manager in valgrind case to catch errors where port bind fails. ( #934 )
2017-09-05 20:58:12 -07:00
Philipp Moritz
7030ef366f
Rebase Ray on latest arrow (remove numbuf from Ray). ( #910 )
...
* remove some stuff
* put get roundtrip working
* fixes
* more fixes
* cleanup
* fix tests
* latest arrow
* fixes
* fix tests
* fix linting
* rebase
* fixes
* fix bug
* bring back libgcc error
* fix linting
* use official arrow repo
* fixes
2017-09-04 22:58:49 -07:00
Eric Liang
a2814567e1
[webui] Quick fix to timeline on task failure ( #930 )
...
* foo
* update
* Move _add_missing_timestamps to task_profiles function.
2017-09-04 22:58:19 -07:00
Eric Liang
63d8d11714
[webui] Checkboxes should go to the left of their labels ( #932 )
2017-09-04 17:05:13 -07:00
Robert Nishihara
8ed03b1cf0
Make task timeline work with ipywidgets==7.0.0, change slider default values. ( #925 )
...
* Make task timeline work with ipywidgets==7.0.0.
* Change initial UI slider values from 70-100 to 0-100.
2017-09-03 23:15:46 -07:00
Eric Liang
246be812f0
upd ( #917 )
2017-09-02 23:55:10 -07:00
Eric Liang
1ebfe9608f
[rllib] Add downscale and frameskip options for Montezumas ( #908 )
...
* up
* update
* fix
* update
* update
* update
* api break
* Update run_multi_node_tests.sh
* fix
2017-09-02 17:20:56 -07:00
Robert Nishihara
deca29a7eb
Bump version to 0.2.0. ( #877 )
2017-08-29 21:38:35 -07:00
Philipp Moritz
164a8f368e
[rllib] Rename algorithms ( #890 )
...
* rename algorithms
* fix
* fix jenkins test
* fix documentation
* fix
2017-08-29 16:56:42 -07:00
Robert Nishihara
e1831792f8
For PPO, rename num_agents -> num_workers. ( #882 )
2017-08-28 23:11:06 -07:00
Robert Nishihara
1afc487baf
In setup.py, move cython to setup_requires. ( #878 )
...
* In setup.py, move cython to setup_requires and move setuptools_scm to setup_requires.
* Add back pip install of cython when building mac wheels.
* Revert changes to setuptools_scm.
* Check that the correct number of Linux wheels are produced.
* Add back pip install cython when building linux wheels.
2017-08-28 23:07:33 -07:00
Robert Nishihara
60d4d01d06
Use observation filter in compute_action for PPO. ( #884 )
2017-08-28 23:01:29 -07:00
Richard Liaw
5d72818ddc
Generic shared_model class ( #880 )
...
Changing `shared_model` class back to `get_model` rather than `ConvolutionalNetwork`
2017-08-28 22:48:07 -07:00
