Commit Graph

  • cc8f7db246 [docs] Improve cluster/docker docs (#3517) Richard Liaw 2018-12-12 10:40:54 -08:00
  • 5f4a9cc713 [rllib] Rollout should preprocess observations; some cleanups (#3512) Eric Liang 2018-12-11 20:16:38 -08:00
  • 59f4743f20 [rllib] Run simple regressions tests for all algs in jenkins (#3498) Eric Liang 2018-12-11 17:21:53 -08:00
  • e0fbb68e47 [tune] Custom Logging, Trial Name (#3465) Richard Liaw 2018-12-11 13:41:59 -08:00
  • 74c3370bd5 Show slowest tests in travis. (#3507) Robert Nishihara 2018-12-11 14:25:04 -05:00
  • 52df4dfc6f [rllib] Fix multiagent_two_trainer test (#3509) Eric Liang 2018-12-11 00:16:39 -08:00
  • 1f4a01cff6 [tune] Fix PyTorch example after PyTorch v1 (#3500) Richard Liaw 2018-12-10 12:00:53 -08:00
  • 962f18756b [autoscaler] Use fixed timestamp to check against health timeouts (#3503) Eric Liang 2018-12-10 11:58:27 -08:00
  • abd781d607 Make stress test time shorter. (#3506) Yuhong Guo 2018-12-11 03:46:40 +08:00
  • ce388a45cf [rllib] Learner should not see clipped actions (#3496) Eric Liang 2018-12-09 21:57:11 -08:00
  • 87c0d24579 [sgd] Add file lock to protect compilation of sgd op (#3486) Philipp Moritz 2018-12-09 13:52:40 -08:00
  • cffe8f9806 Add option to evict keys LRU from the sharded redis tables (#3499) Eric Liang 2018-12-09 05:48:52 -08:00
  • 0136af5aac Add return value for recontruction RPC. (#3493) Yuhong Guo 2018-12-09 16:08:44 +08:00
  • 7aec357501 [rllib] Multi-GPU support for Multi-Agent PPO (#3479) Eric Liang 2018-12-08 18:02:33 -08:00
  • 8b5827b9da [rllib] Better document which methods are abstract and which ones are overrides (#3480) Eric Liang 2018-12-08 16:28:58 -08:00
  • 462e6ef066 [rllib] Use smoothed version of collect metrics for DQN (#3491) Eric Liang 2018-12-07 18:36:23 -08:00
  • f6490f9bef Resolve no handlers could be found for logger 'ray.worker' when importing ray (#3483) Tianming Xu 2018-12-07 12:46:53 +08:00
  • 8395523f81 [rllib] Copy data before passing to Ape-X learner thread (fixes transient plasma crashes) (#3484) Eric Liang 2018-12-06 18:01:11 -08:00
  • c2c501bbe6 Experimental asyncio support (#2015) Si-Yuan 2018-12-06 17:39:05 -08:00
  • 970babf31a Removing the check about the size re: ray-project/ray#3450 (#3464) Devin Petersohn 2018-12-06 16:59:24 -08:00
  • 7a7c6e53c8 [tune/rllib] Use cloudpickle to dump config (#3462) Eugene Vinitsky 2018-12-06 15:52:44 -08:00
  • b9e1977fae Fix failure of test_free_objects_multi_node (#3481) Yuhong Guo 2018-12-07 04:55:49 +08:00
  • 412aaa5195 [tune] Deprecate ambiguous function values (use tune.function / tune.sample_from instead) (#3457) Eric Liang 2018-12-06 11:35:20 -08:00
  • d864f299d7 [rllib] fixes from dogfooding multi-agent (#3456) Eric Liang 2018-12-05 23:31:45 -08:00
  • 7a79b7f62c increase container memory and shm to 20G (#3475) shane 2018-12-05 14:59:07 -08:00
  • 2e6f9bedf2 Add the extra fallback for serialization (#3468) Si-Yuan 2018-12-05 13:09:08 -08:00
  • 06f6431765 Make test_actor_multiple_gpus_from_multiple_tasks less stressful in travis Philipp Moritz 2018-12-04 17:44:33 -08:00
  • 93a9d32288 [docs] Switch docs to use rllib train instead of train.py Eric Liang 2018-12-04 17:36:06 -08:00
  • 9d0bd50e78 [tune] Component notification on node failure + Tests (#3414) Richard Liaw 2018-12-04 14:47:31 -08:00
  • ce355d13d4 [rllib] Allow envs to be auto-registered; add on_train_result callback with curriculum example (#3451) Eric Liang 2018-12-03 23:15:43 -08:00
  • be6567e6fd Tweak/exec attach info (#3447) Kristian Hartikainen 2018-12-03 21:39:43 -08:00
  • d8205976e8 [rllib] Auto clip actions to Box space range; deprecate squash_to_range (#3426) Eric Liang 2018-12-03 19:55:25 -08:00
  • 7abfbfd2f7 [rllib] Better error message for unsupported non-atari image observation sizes (#3444) Eric Liang 2018-12-03 01:24:36 -08:00
  • 4abafd7e62 Fix bug in ray.wait (#3445) Stephanie Wang 2018-12-01 19:40:33 -08:00
  • 13c8ce4d84 Update README.rst with 0.6.0 version number. (#3453) Eric Liang 2018-12-01 19:16:45 -08:00
  • c5b5cdae33 Upgrade Arrow to include Plasma TensorFlow Op release fix (#3448) Philipp Moritz 2018-12-02 01:15:09 +01:00
  • abd37df41e Add stress test for Java worker (#3424) Hao Chen 2018-12-02 08:11:09 +08:00
  • 0603e0b73a Bump version from 0.5.3 to 0.6.0. (#3420) ray-0.6.0 Robert Nishihara 2018-12-01 11:39:36 -08:00
  • 57512616e1 Update readme to contain logo (#3443) Devin Petersohn 2018-11-30 18:28:35 -08:00
  • 454d3aa07d [docs] Snippet did not have a code-block tag above it (#3442) GiliR4t1qbit 2018-11-30 16:39:40 -08:00
  • 447604a9fe Use actor ID for the dummy object (#3437) Stephanie Wang 2018-11-29 22:31:04 -08:00
  • 07d8cbf414 [rllib] Support batch norm layers (#3369) Eric Liang 2018-11-29 13:33:39 -08:00
  • 4d2010a852 Ship Modin with Ray. (#3109) Devin Petersohn 2018-11-29 11:05:24 -08:00
  • 48a5935224 Fault tolerance for actor creation (#3422) Stephanie Wang 2018-11-29 10:48:35 -08:00
  • fd7e494344 Remove: duplicate feed_dict constructing (#3431) Chunyang Wen 2018-11-30 02:21:46 +08:00
  • 7e319dbf0c Automatically indent tune logger params (#3399) Kristian Hartikainen 2018-11-29 00:15:50 -08:00
  • c46ea2ff4b Click 0.7 changes the naming convention for commands; fix this Eric Liang 2018-11-28 14:59:58 -08:00
  • 139fbf7884 Initialize client_id_ in ObjectManager constructor that takes user-defined ObjectDirectory (#3403) Tianming Xu 2018-11-28 15:51:18 +08:00
  • 82863b5251 [autoscaler] Update autoscaler to use heartbeat batches. (#3409) Robert Nishihara 2018-11-27 23:46:27 -08:00
  • f0df97db6f [rllib] example and docs on how to use parametric actions with DQN / PG algorithms (#3384) Eric Liang 2018-11-27 23:35:19 -08:00
  • c2108ca64f Don't put entire actor registry in debug string since it's too long (#3395) Eric Liang 2018-11-27 16:48:12 -08:00
  • 0d56fc10cc Move setproctitle to ray[debug] package (#3415) Eric Liang 2018-11-27 09:50:59 -08:00
  • 20b8b1d891 Add script for running stress tests. (#3378) Robert Nishihara 2018-11-27 04:28:02 -08:00
  • e3c088fa1e [rllib] PPO doesn't work with fractional num gpus (#3396) Eric Liang 2018-11-27 01:14:10 -08:00
  • aa94d3dd50 [autoscaler] Allow more than 5s from node creation to first heartbeat (#3385) Eric Liang 2018-11-26 17:25:05 -08:00
  • 0f0099fb90 UI changes, fix the task timeline and add the object transfer timeline to UI. (#3397) Robert Nishihara 2018-11-25 10:16:49 -08:00
  • b85e7b43f3 [rllib] Refactor the sampler (#3387) Eric Liang 2018-11-24 18:16:54 -08:00
  • 3856533065 Fix incompatibility with most recent version of Redis. (#3379) Robert Nishihara 2018-11-24 16:36:38 -08:00
  • 18a8dbfcfb [rllib] Clip DDPG ou-noise to avoid exceeding action bounds (#3386) Eric Liang 2018-11-24 00:56:50 -08:00
  • 55fca828ce [rllib] Fix use_lstm option when using custom model with dict space (#3368) Eric Liang 2018-11-23 22:51:08 -08:00
  • 8b76bab25c [rllib] docs for td3 (#3381) Eric Liang 2018-11-22 13:36:47 -08:00
  • 41b6b50d09 fix py3 (#3382) Eric Liang 2018-11-22 11:43:52 -08:00
  • b9ae5edf74 When getting a role/profile, catch only exception that indicates the role/profile already exists, allow others to be raised (#3383) GiliR4t1qbit 2018-11-22 09:42:58 -08:00
  • 24bfe8ab76 Enable Twin Delayed DDPG for RLlib DDPG agent (#3353) Jones Wong 2018-11-22 12:03:20 +08:00
  • 6b3236349c Fix memory leak in lineage cache (#3366) Stephanie Wang 2018-11-21 16:18:39 -08:00
  • 784a6399b0 [tune] Node Fault Tolerance (#3238) Richard Liaw 2018-11-21 12:38:16 -08:00
  • 3e33f6f71b Fix failure handling for actor death (#3359) Stephanie Wang 2018-11-21 12:26:22 -08:00
  • 1a926c9b7c Fix $MACOSX_DEPLOYMENT_TARGET (#3337) Philipp Moritz 2018-11-21 10:56:17 -08:00
  • 686cf20951 Remove uses of std::list::size (#3358) Eric Liang 2018-11-20 14:47:55 -08:00
  • c24d87b4d1 [autoscaler] Submit command (#3312) Richard Liaw 2018-11-20 14:03:34 -08:00
  • d3697ce4e1 Ready queue refactor to make Dispatching tasks more efficient (#3324) Philipp Moritz 2018-11-20 13:14:12 -08:00
  • b0bfd104f2 Batch heartbeats from node manager together in the monitor. (#3011) Ujval Misra 2018-11-20 09:52:27 -08:00
  • abdc3b592e [rllib] Update multi-gpu impala numbers (#3327) Eric Liang 2018-11-19 20:55:27 -08:00
  • 5972c29d28 [rllib] Set ape-x local exploration to 0, also load explorations before training steps (#3349) Eric Liang 2018-11-19 20:36:25 -08:00
  • afc48d7b77 Don't setpgid() on actors (#3347) Eric Liang 2018-11-19 17:35:26 -08:00
  • f2b5500642 Add ordered_set container. (#3352) Robert Nishihara 2018-11-19 17:01:18 -08:00
  • d4dbd27e0d Don't retry IPC connect an absurd number of times (#3355) Eric Liang 2018-11-19 16:23:59 -08:00
  • e4bb5d8d16 Fix logging when ray cluster utils is used Eric Liang 2018-11-18 21:49:27 -08:00
  • 61e3bbbfee Update stale example links Eric Liang 2018-11-17 15:40:38 -08:00
  • 5cbc597494 Suppress duplicate pre-emptive object pushes. (#3276) Robert Nishihara 2018-11-16 23:02:45 -08:00
  • ab1e0f5c2f support home path and relative path for temp-dir (#3329) Wenting Shen 2018-11-17 09:41:10 +08:00
  • 60b22d9a72 Don't unsubscribe dependencies for infeasible tasks. (#3338) Robert Nishihara 2018-11-16 11:33:00 -08:00
  • e0bf9d7305 Add debug string to raylet (#3317) Eric Liang 2018-11-15 21:47:50 -08:00
  • d10cb570ab Rename _submit -> _remote. (#3321) Robert Nishihara 2018-11-15 15:30:18 -08:00
  • 98edf752a9 Note requirement cython==0.27.3 in installation instructions. (#3322) Robert Nishihara 2018-11-15 15:27:19 -08:00
  • 1be1455d86 Fix redis crash when duplicate messages are appended to log. (#3316) Philipp Moritz 2018-11-15 15:09:39 -08:00
  • 5723291db6 Raise exception if the node is nearly out of memory (#3323) Eric Liang 2018-11-15 12:55:25 -08:00
  • b6a12d1f97 Fix socket retry message (#3325) Philipp Moritz 2018-11-15 12:14:19 -08:00
  • 5319fd044c Update redis version in setup.py (#3333) Lewis Belcher 2018-11-15 19:40:08 +01:00
  • 706dc1d473 [rllib] Add test for multi-agent support and fix IMPALA multi-agent (#3289) Eric Liang 2018-11-14 14:14:07 -08:00
  • 57c7b4238e KL Divergence Metrics (#3300) andrewztan 2018-11-13 23:12:35 -08:00
  • 1660c9d627 Kill actor child processes on shutdown (#3297) Eric Liang 2018-11-13 19:16:42 -08:00
  • 577c1dda74 Release sender connections as soon as WriteMessageAsync completes (#3313) Stephanie Wang 2018-11-13 18:32:24 -08:00
  • 9d4847ad2d [hot-fix] Fix error when calling Ray.init() twice. (#3314) Wang Qing 2018-11-14 10:21:54 +08:00
  • 65c27c70cf [rllib] Clean up agent resource configurations (#3296) Eric Liang 2018-11-13 18:00:03 -08:00
  • d4fad222e1 Update profiling instructions for raylet (#3311) Philipp Moritz 2018-11-13 14:48:33 -08:00
  • 97f423781b Clean up Ray processes after cluster util exits (#3278) Richard Liaw 2018-11-13 13:18:12 -08:00
  • c3a2c7ebed [tune] Doc: Autofilled, StatusReporter (#3294) Richard Liaw 2018-11-13 13:15:56 -08:00
  • 6ee7a3b571 [rllib] Raise worker TF intra_op threads to 2, lower driver intra_op threads to 8 (#3299) Eric Liang 2018-11-13 11:41:58 -08:00
  • c0423db05c [core] Add Global State Test for multi-node setting (#3239) Richard Liaw 2018-11-13 10:35:24 -08:00