Commit graph

2764 commits

Author SHA1 Message Date
Eric Liang
4f46d3e9bf
[rllib] Add multi-agent examples for hand-coded policy, centralized VF (#4554) 2019-04-09 00:36:49 -07:00
Duane
7f23e8431b fixed typo in kuber yaml (#4582) 2019-04-08 23:13:42 -07:00
Stefan Pantic
915486984a [autoscaler] Add support for separate docker containers on head and worker nodes (#4537)
* Added support for running different docker containers on clusters

* Remove node specific container names

* Keep old options and expand with node specific configuration

* Optimized imports

* Changed docker fields for autoscaler

* Auto reformat

* Updated comments

* Updated condition

* Run linter

* Updated example

* Changed condition for docker images, updated examples

* Removed duplicate line

* Fixed setup_commands

* Update autoscaler.py

* fix_better_image
2019-04-07 16:51:32 -07:00
Jones Wong
da5a471485 [rllib] validate observation in NoPreprocessor (#4546) 2019-04-07 16:11:50 -07:00
Eric Liang
f9b8e77e3b
[rllib] Don't merge unrolls from same episode when calculating seq lens (#4557) 2019-04-07 12:11:30 -07:00
Eric Liang
37208216ae
[rllib] Rename Agent to Trainer (#4556) 2019-04-07 00:36:18 -07:00
Dušan Josipović
820c71b7d0 [tune/rllib] Add checkpoint eraser (#4490) 2019-04-06 20:01:54 -07:00
ctombumila37
7746d20d30 [rllib] ExternalMultiAgentEnv (#4200) 2019-04-06 19:58:14 -07:00
Andrew Tan
991b911e1d [tune] Add --columns flag for CLI (#4564) 2019-04-05 19:49:01 -07:00
Jérémy
300ec72d15 [tune] Add compatibility to nevergrad 0.2.0+ (#4529)
## What do these changes do?

This PR prepares for future version  0.2.0 of `nevergrad`, in which each suggestion is a `Candidate` instance having fields `args` and `kwargs` instead of being a `np.ndarray`. The proposed changes are compatible with all versions of `nevergrad` (manually tested with `nevergrad_example.py` on both `master` and current version `v0.1.6`).

See `nevergrad`'s [CHANGELOG](https://github.com/facebookresearch/nevergrad/blob/master/CHANGELOG.md) for more information on the change.

## Related issue number

None

## Linter

- [x] I've run `scripts/format.sh` to lint the changes in this PR.
2019-04-05 19:44:58 -07:00
Andrew Tan
bfd0af52bc [tune] Add documentation to --output flag (#4518)
## What do these changes do?

Add documentation for the `--output` flag for ls / lsx in the Tune CLI.

## Related issue number

Closes #4511 

## Linter

- [x] I've run `scripts/format.sh` to lint the changes in this PR.
2019-04-05 00:16:35 -07:00
Richard Liaw
50b2aa0740
[tune] Better handling of tune.function in global checkpoint (#4519)
Enables result keys to be queried by CLI.
2019-04-04 21:08:47 -07:00
Federico Fontana
fb88f7efe6 Fixed bug in Dirichlet (#4440) (#4560) 2019-04-04 14:33:09 -07:00
Tasha Chin
5693cd1344 [docs] Show source code (#3281) 2019-04-03 21:30:20 -07:00
William Ma
4b25810994 Adds a push_id to every push in the object manager (#4407) 2019-04-03 17:12:06 -07:00
Yuhong Guo
c2349cf12d Remove local/global_scheduler from code and doc. (#4549) 2019-04-03 17:05:09 -07:00
Adi Zimmerman
51dae23d5c [tune] Search Alg delay import + CLI timing test (#4230) 2019-04-03 08:52:45 -07:00
cclauss
68ccc4d3cf Travis CI: Do not hard-code Trusty, it EOLs this month (#4545)
* Travis CI: Do not hard-code Trusty, it EOLs this month

Do not hard-code __Trusty__ because it reaches its end-of-life this month.
https://wiki.ubuntu.com/Releases

* Update .travis.yml
2019-04-03 16:39:52 +08:00
Philipp Moritz
b0f6ddf6d1 Remove CMake files (#4493) 2019-04-02 22:17:33 -07:00
Wang Qing
7d776f35e1 Integrate metrics (#4246) 2019-04-02 21:01:02 -07:00
cclauss
8e19d3721f Travis CI: The 'sudo' tag is now deprecated (#4542)
[Travis are now recommending removing the __sudo__ tag](https://blog.travis-ci.com/2018-11-19-required-linux-infrastructure-migration).

"_If you currently specify __sudo: false__ in your __.travis.yml__, we recommend removing that configuration_"
2019-04-03 10:55:19 +08:00
Hao Chen
23404f7bcf Fix some flaky tests (#4535) 2019-04-02 17:57:11 -07:00
Simon Mo
db4cf24636 [serve] Double Serialization Optimization (#4532) 2019-04-02 12:35:03 -07:00
Eric Liang
55a2d39409
[rllib] Add option for RNN state and value estimates to span episodes (#4429)
* wip soft horizon

* tests
2019-04-02 02:44:15 -07:00
Yuhong Guo
c2c548bdfd Fix broken pipe callback (#4513) 2019-04-02 17:42:18 +08:00
bibabolynn
20c7b2a6eb [Java] TestNG outputs more verbose error messages (#4507)
[Java] TestNG outputs more verbose error messages
2019-04-02 17:41:20 +08:00
Jones Wong
fe7763e786 [rllib] replace the assertion in SyncReplayOptimizer by a warning (#4534) 2019-04-02 01:43:22 -07:00
opherlieber
60b230b8ad [rllib] Add support for LR schedule to DQN/APEX (#4473) 2019-04-01 11:35:34 -07:00
Eric Liang
0d94f3eeef
[rllib] Improve datapath throughput of IMPALA / APPO (#4324) 2019-03-31 12:25:52 -07:00
Toanngo
dffe19c59c Update GCP gpu image (#4524) 2019-03-31 01:01:22 -07:00
Eric Liang
b01ac41e6f
[rllib] Try to call close on envs on stop (#4521) 2019-03-30 19:36:05 -07:00
Eric Liang
fce0062380
[rllib] Switch to tune.run() instead of run_experiments() (#4515) 2019-03-30 14:07:50 -07:00
ppeagle
5efb21e1d0 Initial commit for Ray streaming (#4268) 2019-03-30 19:32:05 +08:00
Eric Liang
e5bcae52f5
Add lint advisory to PR template 2019-03-29 16:49:02 -07:00
Risto Vuorio
798944fbfa Fixes Inconsistent weight assignment operations in DQNPolicyGraph (#4… (#4504)
* Fixes Inconsistent weight assignment operations in DQNPolicyGraph (#4502)

* Update dqn_policy_graph.py
2019-03-29 13:31:59 -07:00
Leon Sievers
f4b313eaad [rllib] Moved clip_action into policy_graph; Clip actions in compute_single_action (#4459)
* Moved clip_action into policy_graph; Clip actions in compute_single_action

* Update policy_graph.py

* Changed formatting

* Updated codebase for convencience
2019-03-29 13:26:07 -07:00
gehring
5133b10700 Add support for tensorflow resource variables (#4438)
* Adding support for resource variables

Currently resource variable go undetected by the `TensorFlowVariables` since they do not use the same ops for reading values. This change should fix this until a more robust solution is implemented.

* fix varhandle
2019-03-29 13:23:05 -07:00
bjg2
77005d1814 [rllib] Make batch timeout for remote workers tunable (#4435) 2019-03-29 13:19:42 -07:00
Eric Liang
2ffe67c5c3
[rllib] Minor cleanups to TFPolicyGraph: add init args, constants for loss inputs (#4478) 2019-03-29 12:44:23 -07:00
bibabolynn
ab55a1f93a [Java] Clean up outdated dependencies (#4489) 2019-03-28 14:33:45 +08:00
Philipp Moritz
1bcb0b94cc Synchronize arrow version and put changes upstream (#4385) 2019-03-27 22:37:07 -07:00
Robert Nishihara
a22bf1e511 Minor aesthetic changes to python file. (#4492) 2019-03-28 10:59:29 +08:00
Eric Liang
09b2961750
[rllib] Ensure stats are consistently reported across all algos (#4445) 2019-03-27 15:40:15 -07:00
Eric Liang
2871609296
[rllib] Report sampler performance metrics (#4427) 2019-03-27 13:24:23 -07:00
Andrew Tan
12db684f72 [tune] add filter flag for Tune CLI (#4337)
## What do these changes do?

Adds filter flag (--filter) to ls / lsx commands for Tune CLI.

Usage: `tune ls [path] --filter [column] [operator] [value]`
e.g. `tune lsx ~/ray_results/my_project --filter total_trials == 1`
2019-03-27 11:19:25 -07:00
Robert Nishihara
c6f12e5219 Update documentation from 0.7.0.dev1 to 0.7.0.dev2. (#4485) 2019-03-26 17:32:53 -07:00
Robert Nishihara
c0e10ef12d Bump version number from 0.6.5 to 0.7.0.dev2. (#4484) 2019-03-26 16:44:32 -07:00
Robert Nishihara
8548f12eb2 Give better error when include_webui=1 and webui can't be started. (#4471) 2019-03-26 14:54:32 -07:00
bibabolynn
7a9d1546d4 [java] Fix getWorker and add create multi actors test (#4472) 2019-03-26 20:26:13 +08:00
Wang Qing
7d70cfba6e [Java] Fix loading custom classes from jars (#4475) 2019-03-26 20:15:08 +08:00