145 Commits

Author SHA1 Message Date
Nan Zhu
fc85f776f4
allow not stop process in error (#97)
* allow not stop process in error

* fix merge error
2019-06-25 13:04:39 -07:00
Nan Zhu
a429748e24
allow multi call on init (#92) 2019-04-26 18:41:02 -07:00
Chen Qin
5c3b36f346 Allow using external dmlc-core (#91)
* Set `RABIT_BUILD_DMLC=1` if use dmlc-core in rabit

* remove dmlc-core
2019-04-26 15:28:45 +08:00
Chen Qin
e3d51d3e62 [rabit harden] Enable all tests (#90)
* include osx in tests
* address `time_wait` on port assignment
* increase submit attempts.
* cleanup tests
2019-04-24 19:12:11 +08:00
Chen Qin
ecd4bf7aae [rabit harden] replace hardcopy dmlc-core headers with submodule links (#86)
* backport dmlc header changes to rabit

* use gitmodule to reference latest dmlc header files

* include ref to dmlc-core
fix cmake

* update cmake file, add cmake build traivs task

* try force using g++-4.8

* per feedback, update cmake
2019-03-23 13:11:29 +08:00
Chen Qin
ed06e0c6af [rabit harden] fix rabit tests (#81)
* enable model recovery tests
* force use gcc4.8 in Travis
2019-03-15 07:16:45 +08:00
Jiaming Yuan
1cc34f01db
Fix ssize_t definition. (#80)
* Fix linter.
2019-02-18 19:25:08 +08:00
Jiaming Yuan
05941a5f96
Try fixing mingw build error when using CMake. (#77)
* Try fixing mingw build error when using CMake.

* Check __MINGW32__ .

* Fix linter.
2019-02-16 22:35:43 +08:00
Chen Qin
eb2590b774 workaround macosx java test race condition (#74)
* fix error in dmlc#57, clean up comments and naming

* include missing packages, disable recovery tests for now

* disable local_recover tests until we have a bug fix

* support larger cluster

* fix lint, merge with master

* fix mac osx test failure in https://github.com/dmlc/xgboost/pull/3818

* Update allreduce_robust.cc
2018-10-26 12:39:31 -07:00
Chen Qin
3a35dabfae support larger cluster (#73)
* fix error in dmlc#57, clean up comments and naming

* include missing packages, disable recovery tests for now

* disable local_recover tests until we have a bug fix

* support larger cluster

* fix lint, merge with master
2018-10-22 10:13:45 -07:00
AbdealiJK
21b5e12913 allreduce_robust.cc: Allow num_global_replica to be 0 (#38)
In some cases, users may not want to have any global replica of
the data being broadcasted/all-reduced. In such cases, set the
result_buffer_round to -1 as a flag that this is not necessary
and check for it.
2016-11-23 19:34:11 -08:00
tomlaube
1007a26641 Fixing the imports to work with MPI (#30) 2016-08-26 15:04:41 -07:00
Johnny Ho
9d235c31a7 Use getaddrinfo instead of gethostbyname for thread safety 2016-07-27 02:35:02 -04:00
tqchen
e19fced5cb [FIX] rabit on single node 2016-05-10 20:05:59 -07:00
tqchen
849b20b7c8 add distributed checking 2016-04-11 15:43:01 -07:00
tqchen
be50e7b632 Make rabit library thread local 2016-03-01 20:12:51 -08:00
tqchen
aeb4008606 remove connect msg 2016-02-29 16:27:48 -08:00
tqchen
c7d53aecc3 add link tag 2016-02-28 09:44:11 -08:00
tqchen
f0f07ecd22 fix 2016-02-27 20:51:00 -08:00
tqchen
7479791f6a refactor: librabit 2016-02-27 10:14:41 -08:00
tqchen
112d866dc9 [RABIT] fix rabit in local mode 2016-01-12 21:34:26 -08:00
Andrey
291ab05023 Remove redundant whitespace again 2015-11-18 10:21:03 +03:00
Andrey
de251635b1 Remove redundant whitespace 2015-11-18 00:53:53 +03:00
Andrey
3a6be65a20 Fix bug with name of sleep function 2015-11-17 21:45:52 +03:00
yonglong.dyl
35c3b371ea add retry mechanism to ConnectTracker and modify Listen backlog to 128
in rabit_traker.py
2015-10-21 10:24:07 +08:00
tqchen
3cc49ad0e8 lint and travis 2015-07-03 15:15:11 -07:00
tqchen
ceedf4ea96 fix 2015-05-28 12:37:06 -07:00
Tianqi Chen
fd8920c71d fix win32 2015-05-28 12:24:26 -07:00
Chuntao Hong
df14bb1671 fix type 2015-05-20 11:36:17 +08:00
Chuntao Hong
f441dc7ed8 replace tab with blankspace 2015-05-20 11:33:48 +08:00
Chuntao Hong
181ef47053 defined long long and ulonglong 2015-05-20 11:27:50 +08:00
tqchen
e0b7da0302 fix 2015-05-02 21:47:43 -07:00
tqchen
fa99857467 try fix warning on some platforms 2015-05-01 22:45:11 -07:00
tqchen
50a66b3855 fix empty engine 2015-04-09 08:44:33 -07:00
tqchen
e95c96232a remove I prefix from interface, serializable now takes in pointer 2015-04-08 15:25:58 -07:00
tqchen
8cb4c02165 add dmlc support 2015-03-28 22:44:10 -07:00
tqchen
146e069000 bugfix: logical boundary for ring buffer 2015-03-11 20:28:34 -07:00
tqchen
4cf3c13750 Merge branch 'master' of ssh://github.com/tqchen/rabit
Conflicts:
	tracker/rabit_tracker.py
2015-03-11 13:35:35 -07:00
tqchen
c57dad8b17 add ringbased passing and batch schedule 2015-03-11 12:00:19 -07:00
tqchen
091634b259 fix 2015-03-09 14:56:01 -07:00
tqchen
d558f6f550 redefine distributed means 2015-03-09 14:43:05 -07:00
tqchen
67ebf81e7a allow setup from env variables 2015-03-07 16:45:31 -08:00
tqchen
d4ec037f2e fix rabit 2015-03-03 13:12:05 -08:00
tqchen
d29892cb22 add mock option statis 2015-03-02 16:10:08 -08:00
tqchen
4db0a62a06 bugfix of lazy prepare 2015-02-11 20:31:46 -08:00
tqchen
85b746394e change def of reducer to take function ptr 2015-01-19 21:24:52 -08:00
tqchen
fe6366eb40 add engine base 2015-01-19 19:11:15 -08:00
tqchen
1db6449b01 remove include in -I, make things easier to direct compile 2015-01-18 21:30:19 -08:00
Tianqi Chen
56a80f431b check in windows solutions, pass small test in windows 2015-01-16 20:56:34 -08:00
tqchen
f5245c615c ok 2015-01-16 10:12:47 -08:00