@@ -3,6 +3,49 @@ DGL release and change logs
3
3
4
4
Refer to the roadmap issue for the on-going versions and features.
5
5
6
+ 0.2
7
+ ---
8
+ Major release that includes many features, bugfix and performance improvement.
9
+ Speed of GCN model on Pubmed dataset has been improved by ** 4.19x** ! Speed of
10
+ RGCN model on Mutag dataset has been improved by ** 7.35x** ! Important new
11
+ feature: ** graph sampling APIs** .
12
+
13
+ Update details:
14
+
15
+ # Model examples
16
+ - [x] TreeLSTM w/ MXNet (PR #279 by @szha )
17
+ - [x] GraphSage (@ZiyueHuang )
18
+ - [x] Improve GAT model speed (PR #348 by @jermainewang )
19
+
20
+ # Core system improvement
21
+ - [x] Immutable CSR graph structure (PR #342 by @zheng-da )
22
+ - [x] Finish remaining functionality (Issue #369 , PR #404 by @yzh119 )
23
+ - [x] Nodeflow data structure (PR #361 by @zheng-da )
24
+ - [x] Neighbor sampler (PR #322 )
25
+ - [x] Layer-wise sampler (PR #362 by @GaiYu0 )
26
+ - [x] Multi-GPU support by data parallelism (PR #356 #338 by @ylfdq1118 )
27
+ - [x] More dataset:
28
+ - [x] Reddit dataset loader (PR #372 by @ZiyueHuang )
29
+ - [x] PPI dataset loader (PR #395 by @sufeidechabei )
30
+ - [x] Mini graph classification dataset (PR #364 by @mufeili )
31
+ - [x] NN modules (PR #406 by @jermainewang @mufeili )
32
+ - [x] GraphConv layer
33
+ - [x] Edge softmax layer
34
+ - [x] Edge group apply API (PR #358 by @VoVAllen )
35
+ - [x] Reversed graph and transform.py module (PR #331 by @mufeili )
36
+ - [x] Max readout (PR #341 by @mufeili )
37
+ - [x] Random walk APIs (PR #392 by @BarclayII )
38
+
39
+ # Tutorial/Blog
40
+ - [x] Batched graph classification in DGL (PR #360 by @mufeili )
41
+ - [x] Understanding GAT (@sufeidechabei )
42
+
43
+ # Project improvement
44
+ - [x] Python lint check (PR #330 by @jermainewang )
45
+ - [x] Win CI (PR #324 by @BarclayII )
46
+ - [x] Auto doc build (by @VoVAllen )
47
+ - [x] Unify tests for different backends (PR #333 by @BarclayII )
48
+
6
49
0.1.3
7
50
-----
8
51
Bug fix
0 commit comments