mirrors / dmlc--dgl

文件历史

4,371 提交 34 分支 42 Git 标签
2.4.x
提交图

2073 次代码提交

作者 SHA1 备注 提交日期
Gan Quan 16da76c4b4 fix #81 2018-10-16 10:39:50 -04:00
Da Zheng 5567f4a4d9 [BACKEND] Add MXNet backend. (#77) 
* support mxnet.

* add mxnet version of GCN.

* rename mxnet.nd as F.

* add mxnet GAT.

* enable GPU for GCN.

* fix MXNet GCN train.

* Use adam to optimize GAT

* support more operators.

* support sparse arrays.

* update mxnet backend.

* support index_copy.

* remove NN.

* update mxnet backend.

* temp check in.

* fix data conversion.

* add test.

* clean up mxnet backend.

* update mxnet examples.

* Revert "remove NN."

This reverts commit d815d9a0ec619f9ce9099c48cd35db9d8e947483.

* temp disable MXNet version of NN.
 2018-10-14 11:15:48 -04:00
GaiYu0 e7679cf2e6 implement requested changes 2018-10-13 14:24:46 +08:00
GaiYu0 eb46dc0018 Merge branch 'line-graph' of https://github.com/jermainewang/dgl into line-graph 2018-10-10 12:43:39 +08:00
GaiYu0 0a6d720c30 fix build failure 2018-10-10 12:43:05 +08:00
GaiYu0 4baed9645c Merge branch 'cpp' into line-graph 2018-10-10 07:53:44 +08:00
Minjie Wang 7603b8c351 Fix dlpack bug; zero copy between dgl.ndarray and user tensor enabled 2018-10-09 11:37:59 -04:00
GaiYu0 885f5e7313 Merge branch 'cpp' of https://github.com/jermainewang/dgl into line-graph 2018-10-09 16:08:18 +08:00
Minjie Wang 3d38c2a93e small fix 2018-10-08 13:30:50 -04:00
GaiYu0 baf5906bab (non)-backtracking line graph 2018-10-08 08:04:25 +08:00
GaiYu0 5119a50475 sparse line graph 2018-10-08 07:38:03 +08:00
Lingfan Yu cdf7334cc6 Dockerfile and CI (#74) 
* Test CPP branch CI (#2)

* Fix batching node-only graphs (#62)

* fixing batching with graphs with no edges

* oops forgot test

* fix readme

* Docker and Jenkins (#1)

* docker ci cpu

* install python packages

* docker ci gpu

* add readme

* use dgl cpu image

* run command in container as root

* use python3

* fix test case

* remove nose from docker file

* docker folder readme

* parallelize cpu and gpu

* top level stages

* comment out python2 related installation

* fix

* remove igraph

* building for cpp

* change building order

* export env in test stage

* withEnv

* run docker container as root

* fix test cases

* fix test cases

* minor

* remove old build
 2018-10-07 19:18:43 -04:00
GaiYu0 905db29286 cpp lg 2018-10-07 22:25:45 +08:00
Gan Quan ec4216dda8 [MODEL] junction tree vae example (#70) 
* junction tree vae example

* added README and moved data to dropbox

* auto download; some fix in python3
 2018-10-06 16:59:47 -04:00
GaiYu0 596ca471a5 Merge branch 'cpp' of https://github.com/jermainewang/dgl into line-graph 2018-10-06 23:52:04 +08:00
GaiYu0 8a6d25ae69 Resolve conflicts in python/dgl/graph.py 2018-10-06 16:13:06 +08:00
Minjie Wang 72f6345515 Fix for subgraph test and some docs 2018-10-05 23:58:42 -04:00
Da Zheng b0e02e5b52 Update the subgraph (#73) 
* update subgraph.

* update subgraph API.

* keep node embedding.
 2018-10-05 22:17:33 -04:00
Minjie Wang 1209978d97 install doc 2018-10-05 16:23:55 -04:00
Minjie Wang 1b06998465 doc for batched graph 2018-10-05 15:07:42 -04:00
Minjie Wang c7b9f06946 Init sphinx doc 2018-10-05 14:57:09 -04:00
Minjie Wang c468e06867 some fix 2018-10-05 10:38:44 -04:00
Minjie Wang 74888fee5c readout signature 2018-10-04 13:20:25 -04:00
GaiYu0 26dcfb5bc2 Merge branch 'cpp' of https://github.com/jermainewang/dgl into cpp 
Conflicts:
	python/dgl/graph.py
 2018-10-04 21:26:56 +08:00
GaiYu0 cc372e3774 sbm mixture 2018-10-04 21:22:41 +08:00
Minjie Wang fde4f581d2 gcn runnable 2018-10-03 21:05:19 -04:00
Minjie Wang 2be55fb50a graph batch and unbatch 2018-10-03 17:07:41 -04:00
Minjie Wang 7d04c8c9f3 remove nonbatchable mode 2018-10-03 12:20:12 -04:00
Minjie Wang a8a4fcbaff quickly integrating with tree-lstm example 2018-09-24 23:42:56 -04:00
Minjie Wang 882e2a7be0 Graph batching. Support convert nx graph attrs 2018-09-24 21:34:51 -04:00
Minjie Wang 314a75f367 WIP: batched graph 2018-09-24 17:31:50 -04:00
Minjie Wang 79f62400fe WIP: batch graphs 2018-09-21 14:53:25 -04:00
Minjie Wang 4aebfd7bd8 add merge graph; fix bug in from_networkx 2018-09-21 12:25:54 -04:00
Minjie Wang 7b5d4c5886 pass specialization test 2018-09-20 18:22:46 -04:00
Minjie Wang b2e4bdc03f passed basic test batching 2018-09-20 14:44:06 -04:00
Gan Quan f26ecb4909 Fix batching node-only graphs (#62) 
* fixing batching with graphs with no edges

* oops forgot test
 2018-09-20 10:36:35 -04:00
Minjie Wang 44db98c4e0 remove nx inheritence 2018-09-19 22:49:14 -04:00
Minjie Wang 916d375ba3 Merge branch 'master' into cpp 2018-09-19 21:04:31 -04:00
Minjie Wang a1038eb1ee graph index 2018-09-19 21:02:15 -04:00
Lingfan Yu 9b0a01db41 Support list of msg or reduce func for update_all and send_and_recv (#58) 
* support multiple fields in spmv

* finishing SPMV executor

* non-spmv case

* refactor code to give single mfunc rfunc a shortcut

* two test cases to test multiple msg/red func

* catching cases where msg uses anonymous field

* default ALL for update edge

* more corner case test

* print failed test

* delete print

* fix builtin max reducer
 2018-09-19 20:49:29 -04:00
Minjie Wang a81d27dc67 WIP: integrating 2018-09-19 10:52:20 -04:00
Minjie Wang dba79f1d45 also return edge id for many APIs 2018-09-18 23:40:13 -04:00
Minjie Wang 71e6e0cd5f python shim + some simple tests 2018-09-17 14:40:29 -04:00
Minjie Wang 758cb16e67 ndarray argument 2018-09-17 00:15:46 -04:00
Minjie Wang c086d454d3 Use TVMContext 2018-09-16 16:56:29 -04:00
Minjie Wang b24daa6607 change to rel import within dgl 2018-09-16 16:14:37 -04:00
Minjie Wang 842d37682b python api 2018-09-16 15:49:11 -04:00
Minjie Wang 2694b12725 import ffi solution from TVM 2018-09-05 10:51:31 -04:00
Minjie Wang 61fa3c6cf5 Builtin function and API changes (#53) 
* WIP: API renaming

* API rewrite and node function refactor

* builtin functions

* builtin functions tested

* fix test

* send and recv spmv test

* WIP: fix examples

* Fix examples using new APIs
 2018-09-01 13:38:01 -04:00
Gan Quan b7eb16595f Fix 0deg update_to and Tree-LSTM model (#51) 
* WIP

* WIP

* treelstm dataloader

* Main training loop.

* trainable treelstm script

* fix dependency

* cuda training

* Add tensorized topological traversal

* allowing update_to() with no incoming messages

* fixing partial cases
 2018-08-28 11:57:08 -04:00