Skip to content

Commit aa39ca8

Browse files
authored
Merge pull request #1868 from PaddlePaddle/luotao1-patch-1
Update RELEASE.md
2 parents 0b0d7c0 + d2ddbed commit aa39ca8

File tree

1 file changed

+24
-0
lines changed

1 file changed

+24
-0
lines changed

RELEASE.md

Lines changed: 24 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -1,3 +1,27 @@
1+
# Release v0.10.0
2+
3+
## New Features
4+
5+
* Support rectangle input for CNN.
6+
* Support stride pooling for seqlastin and seqfirstin.
7+
* Expose seq_concat_layer/seq_reshape_layer in `trainer_config_helpers`.
8+
9+
## Improvements
10+
11+
* Speedup the backward of average layer on GPU.
12+
* Reorganize the catalog of doc/ and refine several docs.
13+
* Add Travis-CI for checking dead links.
14+
* Add a example for explaining sparse_vector.
15+
* Add Relu in layer_math.py
16+
17+
## Bug Fixes
18+
19+
* Add layer check for recurrent_group.
20+
* Clang-format off on some cuda .cc files.
21+
* Fix LogActivation which is not defined.
22+
* Fix bug when run test_layerHelpers multiple times.
23+
* Fix protobuf size limit on seq2seq demo.
24+
125
# Release v0.9.0
226

327
## New Features:

0 commit comments

Comments
 (0)