Eric Liang af0c1174cd [sgd] Merge sharded param server based SGD implementation (#3033) ... This includes most of the TF code used for the OSDI experiment. Perf sanity check on p3.16xl instances: Overall scaling looks ok, with the multi-node results within 5% of OSDI final numbers. This seems reasonable given that hugepages are not enabled here, and the param server shards are placed randomly. $ RAY_USE_XRAY=1 ./test_sgd.py --gpu --batch-size=64 --num-workers=N \ --devices-per-worker=M --strategy=<simple|ps> \ --warmup --object-store-memory=10000000000 Images per second total gpus total | simple | ps ======================================== 1 | 218 2 (1 worker) | 388 4 (1 worker) | 759 4 (2 workers) | 176 | 623 8 (1 worker) | 985 8 (2 workers) | 349 | 1031 16 (2 nodes, 2 workers) | 600 | 1661 16 (2 nodes, 4 workers) | 468 | 1712 <--- OSDI perf was 1817		2018-10-27 21:25:02 -07:00
..
_static	Move documentation to ReadTheDocs. (#326)	2017-02-27 21:14:31 -08:00
_templates	Add better analytics to docs (#1854)	2018-04-10 00:51:44 -07:00
images	[tune] Tune Facelift (#2472)	2018-08-19 11:00:55 -07:00
actors.rst	Remove legacy Ray code. (#3121)	2018-10-26 13:36:58 -07:00
apex.png	[rllib] Document "v2" APIs (#2316)	2018-07-01 00:05:08 -07:00
api.rst	Add ray.is_initialized() function. (#2818)	2018-09-06 21:20:59 -07:00
autoscaler-status.png	Update the pip wheel in example.yaml and add docs (#1381)	2018-01-01 13:02:05 -08:00
autoscaling.rst	[autoscaler] Add tmux support for attach and exec (#2907)	2018-09-26 23:22:45 -07:00
conf.py	Remove legacy Ray code. (#3121)	2018-10-26 13:36:58 -07:00
contact.rst	Add mailing list to README and documentation. (#950)	2017-09-09 10:21:51 -07:00
development.rst	Fix yapf excludes, print diff in --all mode (#2211)	2018-06-08 02:25:55 -07:00
es.png	[rllib] Document "v2" APIs (#2316)	2018-07-01 00:05:08 -07:00
example-a3c.rst	[rllib] Propagate model options correctly in ARS / ES, to action dist of PPO (#2974)	2018-10-01 12:49:39 -07:00
example-cython.rst	Replace python setup.py install with pip install -e . (#1460)	2018-02-22 11:15:03 -08:00
example-evolution-strategies.rst	[tune] Support user-defined trainable functions / classes / envs with a shared object registry (#1226)	2017-11-20 17:52:43 -08:00
example-lbfgs.rst	Change Python examples in documentation to use 4 space indentation. (#736)	2017-07-16 22:19:33 -07:00
example-parameter-server.rst	Synchronous parameter server example. (#1220)	2017-11-15 17:49:31 -08:00
example-policy-gradient.rst	[rllib] Propagate model options correctly in ARS / ES, to action dist of PPO (#2974)	2018-10-01 12:49:39 -07:00
example-resnet.rst	Allow ResNet example to run on multiple machines. (#891)	2017-08-29 21:37:53 -07:00
example-rl-pong.rst	Change Python examples in documentation to use 4 space indentation. (#736)	2017-07-16 22:19:33 -07:00
example-streaming.rst	Add streaming MapReduce example (#1251)	2017-11-27 21:38:35 -08:00
fault-tolerance.rst	Remove legacy Ray code. (#3121)	2018-10-26 13:36:58 -07:00
impala.png	[rllib] Basic IMPALA implementation (using deepmind's reference vtrace.py) (#2504)	2018-08-01 20:53:53 -07:00
index.rst	Add password authentication to Redis ports (#2952)	2018-10-16 22:48:30 -07:00
install-on-docker.rst	Fix installation instructions on Ubuntu and convert md -> rst. (#389)	2017-03-24 17:33:26 -07:00
installation-troubleshooting.rst	Rebase Ray on latest arrow (remove numbuf from Ray). (#910)	2017-09-04 22:58:49 -07:00
installation.rst	Add links for latest Python 3.7 wheels to documentation. (#3091)	2018-10-19 12:15:22 -07:00
internals-overview.rst	Remove legacy Ray code. (#3121)	2018-10-26 13:36:58 -07:00
multi-agent.svg	[rllib] Document "v2" APIs (#2316)	2018-07-01 00:05:08 -07:00
pandas_on_ray.rst	Dataframe deprecation (#2353)	2018-07-06 00:16:22 -07:00
pbt.png	[tune] clean up population based training prototype (#1478)	2018-02-02 23:03:12 -08:00
plasma-object-store.rst	Add note about huge pages using up memory. (#2733)	2018-08-24 17:02:54 -07:00
ppo.png	[rllib] Document "v2" APIs (#2316)	2018-07-01 00:05:08 -07:00
profiling.rst	Documentation- Basic Profiling for Ray Users (#2326)	2018-07-12 16:57:39 -07:00
ray-tune-parcoords.png	[tune] Fix Docs (#1469)	2018-01-25 16:39:00 -08:00
ray-tune-tensorboard.png	[tune] Documentation for Ray.tune (#1243)	2017-11-23 11:31:59 -08:00
ray-tune-viskit.png	[tune] Documentation for Ray.tune (#1243)	2017-11-23 11:31:59 -08:00
redis-memory-management.rst	Doc: redis memory management / automatic flushing. (#2344)	2018-07-05 23:44:37 -07:00
resources.rst	Remove dependence on psutil. Add utility functions for getting system memory. (#2892)	2018-09-18 15:03:29 +08:00
rllib-algorithms.rst	[rllib] Include config dicts in the sphinx docs (#3064)	2018-10-16 15:55:11 -07:00
rllib-api.svg	[rllib] Update docs with api and components overview figures (#1443)	2018-01-19 10:08:45 -08:00
rllib-components.svg	[rllib] Update docs with api and components overview figures (#1443)	2018-01-19 10:08:45 -08:00
rllib-concepts.rst	[rllib] Native support for Dict and Tuple spaces; fix Tuple action spaces; add prev a, r to LSTM (#3051)	2018-10-20 15:21:22 -07:00
rllib-config.svg	[rllib] clarify train batch size for PPO (#2793)	2018-09-05 12:06:13 -07:00
rllib-env.rst	[rllib] Propagate model options correctly in ARS / ES, to action dist of PPO (#2974)	2018-10-01 12:49:39 -07:00
rllib-envs.svg	[rllib] Document "v2" APIs (#2316)	2018-07-01 00:05:08 -07:00
rllib-models.rst	[rllib] Native support for Dict and Tuple spaces; fix Tuple action spaces; add prev a, r to LSTM (#3051)	2018-10-20 15:21:22 -07:00
rllib-package-ref.rst	[rllib] Fix atari reward calculations, add LR annealing, explained var stat for A2C / impala (#2700)	2018-08-23 17:49:10 -07:00
rllib-stack.svg	[rllib] Document "v2" APIs (#2316)	2018-07-01 00:05:08 -07:00
rllib-training.rst	[sgd] Merge sharded param server based SGD implementation (#3033)	2018-10-27 21:25:02 -07:00
rllib.rst	[rllib] Include config dicts in the sphinx docs (#3064)	2018-10-16 15:55:11 -07:00
security.rst	Add password authentication to Redis ports (#2952)	2018-10-16 22:48:30 -07:00
serialization.rst	Change Python examples in documentation to use 4 space indentation. (#736)	2017-07-16 22:19:33 -07:00
tempfile.rst	Remove legacy Ray code. (#3121)	2018-10-26 13:36:58 -07:00
throughput.png	[rllib] Document "v2" APIs (#2316)	2018-07-01 00:05:08 -07:00
timeline.png	[minor] Use a better timeline pic in the documentation	2017-12-20 12:54:25 -08:00
troubleshooting.rst	Change Python examples in documentation to use 4 space indentation. (#736)	2017-07-16 22:19:33 -07:00
tune-package-ref.rst	[tune] Tune Facelift (#2472)	2018-08-19 11:00:55 -07:00
tune-schedulers.rst	[tune] Tune Facelift (#2472)	2018-08-19 11:00:55 -07:00
tune-searchalg.rst	[tune] Tweaks to Trainable and Verbosity (#2889)	2018-10-11 23:42:13 -07:00
tune-usage.rst	[tune] Fix SearchAlg finishing early (#3081)	2018-10-22 12:17:13 -07:00
tune.rst	Fixed Typo. (#2865)	2018-09-13 13:32:56 +08:00
tutorial.rst	Remove legacy Ray code. (#3121)	2018-10-26 13:36:58 -07:00
user-profiling-timeline.gif	Documentation- Basic Profiling for Ray Users (#2326)	2018-07-12 16:57:39 -07:00
user-profiling.rst	Documentation- Basic Profiling for Ray Users (#2326)	2018-07-12 16:57:39 -07:00
using-ray-and-docker-on-a-cluster.md	Enable starting and stopping ray with "ray start" and "ray stop". (#628)	2017-06-02 20:17:48 +00:00
using-ray-on-a-cluster.rst	Remove legacy Ray code. (#3121)	2018-10-26 13:36:58 -07:00
using-ray-on-a-large-cluster.rst	Remove legacy Ray code. (#3121)	2018-10-26 13:36:58 -07:00
using-ray-with-gpus.rst	Change Python examples in documentation to use 4 space indentation. (#736)	2017-07-16 22:19:33 -07:00
using-ray-with-tensorflow.rst	[tune] Tune Documentation and expose better API (#1681)	2018-03-19 12:55:10 -07:00
webui.rst	say which port is local and which one is remote (#1591)	2018-02-25 10:19:12 -08:00