Commit Graph

3509 Commits

Author SHA1 Message Date
architkulkarni cc1c2c3dc9 [Serve] Use ServeHandle in HTTP proxy (#12523) 2020-12-28 18:33:42 -08:00
Simon Mo 30c22921d9 [Serve] Implement Graceful Shutdown (#13028) 2020-12-28 17:53:53 -08:00
Lavanya Shukla 350917958c [docs] fix wandb url (#13094) 2020-12-28 17:19:17 -08:00
Eric Liang 836c5d5a91 Deprecate experimental / dynamic resources (#13019) 2020-12-28 11:52:36 -08:00
architkulkarni 9a0218fb89 [Serve] [Doc] Front page update (#13032) 2020-12-28 10:19:36 -08:00
Hao Zhang 18f5743416 [Collective][PR 3.5/6] Send/Recv calls and some initial code for communicator caching (#12935)
* other collectives all work

* auto-linting

* mannual linting #1

* mannual linting 2

* bugfix

* add send/recv point-to-point calls

* add some initial code for communicator caching

* auto linting

* optimize imports

* minor fix

* fix unpassed tests

* support more dtypes

* rerun some distributed tests for send/recv

* linting
2020-12-28 09:48:07 -08:00
Sumanth Ratna b11bd22111 [docs] Fix args + kwargs instead of docstrings (#13068)
* functools wraps

* Fix typo (functoools -> functools)
2020-12-23 19:09:23 -08:00
Edward Oakes 3cc213ddf6 [serve] Centralize HTTP-related logic in HTTPState (#13020) 2020-12-23 18:00:02 -06:00
Alex Wu 8df94e33e0 [Autoscaler] New output log format (#12772) 2020-12-23 12:02:55 -08:00
Antoni Baum a4f2dd2138 [Tune]Add integer loguniform support (#12994)
* Add integer quantization and loguniform support

* Fix hyperopt qloguniform not being np.log'd first

* Add tests, __init__

* Try to fix tests, better exceptions

* Tweak docstrings

* Type checks in SearchSpaceTest

* Update docs

* Lint, tests

* Update doc/source/tune/api_docs/search_space.rst

Co-authored-by: Kai Fricke <krfricke@users.noreply.github.com>

Co-authored-by: Kai Fricke <krfricke@users.noreply.github.com>
2020-12-23 09:27:16 -08:00
Ameer Haj Ali d37e2c3a20 [joblib] Fix flaky joblib test. (#13046) 2020-12-23 10:43:34 -06:00
Barak Michener c4e273920f [ray_client]: Insert decorators into the real ray module to allow for client mode (#13031) 2020-12-22 22:51:45 -08:00
Simon Mo bc68260144 [Serve] Handle Bug Fixes (#12971) 2020-12-22 19:13:16 -08:00
Edward Oakes b52cce6632 [serve] Refactor SystemState into EndpointState and BackendState (#13018) 2020-12-21 20:39:13 -06:00
Eric Liang 8068041006 Don't release resources during plasma fetch (#13025) 2020-12-21 18:32:40 -08:00
Edward Oakes 015a0f9935 [serve] Rename replica_tag -> replica in metrics for consistency (#13022) 2020-12-21 17:19:39 -06:00
Eric Liang 03a5b90ed6 Revert "Revert "Increase the number of unique bits for actors to avoi… (#12990) 2020-12-21 15:16:42 -08:00
architkulkarni 8b4b4bf0a2 [Serve] Migrate from Flask.Request to Starlette Request (#12852) 2020-12-21 15:34:15 -06:00
Hao Zhang 5b48480e29 [Collective][PR 3/6] Other collectives (#12864) 2020-12-21 12:48:00 -08:00
Barak Michener 43b9c7811e [ray_client] add client microbenchmarks (#13007) 2020-12-21 12:17:44 -08:00
Ameer Haj Ali 5e2b850836 [autoscaler] Fixes max_workers bug. (#13008) 2020-12-21 10:30:03 -08:00
Kai Yang 5a6801dde7 [Core] Remove delete_creating_tasks (#12962) 2020-12-22 00:01:27 +08:00
Barak Michener c576f0b073 [ray_client] Implement a gRPC streaming logs API for the client (#13001) 2020-12-20 19:35:34 -08:00
Barak Michener e715ade2d1 Support retrieval of named actor handles (#13000)
Change-Id: I05d31c9c67943d2a0230782cbdaa98341584cbc7
2020-12-20 16:34:50 -08:00
Barak Michener 80f6dd16b2 [ray_client] Implement optional arguments to ray.remote() and f.options() (#12985) 2020-12-20 15:43:48 -08:00
Ameer Haj Ali 11f34f72d8 [autoscaler] Do not count head node with min_workers constraint. (#12980) 2020-12-20 14:54:46 -08:00
Barak Michener 7ab9164f1b [ray_client] Integrate with test_basic, test_basic_2 and test_actor (#12964) 2020-12-20 14:54:18 -08:00
Philipp Moritz bf6577c8f4 Switch debugger to sockets and support unicode (#13004) 2020-12-20 12:10:28 -08:00
Ian Rodney d6e243ad46 [serve] Refactor to full control loop design (#12537) 2020-12-20 13:03:57 -06:00
Richard Liaw 038a50af52 [tune] skopt fix-extra-import (#12970)
Signed-off-by: Richard Liaw <rliaw@berkeley.edu>
2020-12-20 01:01:09 -08:00
Amog Kamsetty 4c63917439 [Queue] Add options and shutdown to Queue (#12932)
Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
2020-12-20 00:42:21 -08:00
Amog Kamsetty 51139ed37c [SGD] Fix process group timeout units (#12477) 2020-12-19 21:46:33 -08:00
Dmitri Gekhtman 4832b39066 Suggest mounting into home. Note non-root user. (#12987) 2020-12-19 16:09:24 -08:00
Eric Liang 64c97d25d3 Enable by default new scheduler (#12735) 2020-12-19 13:22:24 -08:00
Amog Kamsetty 5d3c9c8861 [Tune] Mlflow Integration (#12840)
Co-authored-by: Kai Fricke <krfricke@users.noreply.github.com>
Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
2020-12-19 00:40:02 -08:00
Eric Liang 5d987f5988 Revert "Increase the number of unique bits for actors to avoid handle collisions (#12894)" (#12988)
This reverts commit 3e492a79ec.
2020-12-18 23:51:44 -08:00
SangBin Cho 9d939e6674 [Object Spilling] Implement level triggered logic to make streaming shuffle work + additional cleanup (#12773) 2020-12-18 19:31:14 -08:00
Alex Wu 404161a3ff [Autoscaler/Core] Remove autoscaler spam (#12952) 2020-12-18 18:22:45 -08:00
Kai Yang ac5ea2c13d [Java] Fix output parsing in RunManager (#12968)
* Fix output parsing in RunManager

* change log level

Co-authored-by: 灵洵 <fengbin.ffb@antgroup.com>
2020-12-18 18:22:12 -08:00
Eric Liang 6ece291f35 Clean up block/unblock handling of resources in new scheduler (#12963) 2020-12-18 16:00:54 -08:00
Eric Liang 3e492a79ec Increase the number of unique bits for actors to avoid handle collisions (#12894) 2020-12-18 15:59:03 -08:00
Edward Oakes 3521e74f3a [serve] Support for imported backends (#12923) 2020-12-18 15:49:24 -06:00
Eric Liang 92812f2e8a Implement resource deadlock detection for new scheduler (#12961) 2020-12-18 12:17:54 -08:00
Barak Michener 5cfa1934e4 [ray_client]: Implement object retain/release and Data Streaming API (#12818) 2020-12-18 11:47:38 -08:00
Kai Fricke 55ae567f7a [tune] Fix and enable SigOpt tests (#12877)
Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
2020-12-18 01:33:12 -08:00
Gekho457 bff50cfc37 [k8s] Read gpu resources properly (#12942)
* Read gpu resources properly

* Comments and docstrings

* Comment formatting
2020-12-18 01:32:12 -08:00
Kai Fricke 426f8a8d15 [tune] Fix tutorial training on GPU (#12914) 2020-12-18 01:31:40 -08:00
DK.Pino 6404f1e609 [Placement Group][New scheduler] New scheduler pg implementation (#12910) 2020-12-18 11:56:45 +08:00
Farzan Taj 53378170e0 [tune] Change pickle to ray.cloudpickle -- support large models (#12958)
* Change pickle to ray.cloudpickle

* Change pickle import to ray.cloudpickle
2020-12-17 19:17:08 -08:00
Kai Fricke 3d72000826 [tune] Add points_to_evaluate to BasicVariantGenerator (#12916)
Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
2020-12-17 19:16:03 -08:00