Skip to content

Commit 9935d7d

Browse files
committed
Add Release notes
1 parent 341688b commit 9935d7d

File tree

1 file changed

+70
-0
lines changed

1 file changed

+70
-0
lines changed

RELEASE.md

Lines changed: 70 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,70 @@
1+
# Release v0.9.0
2+
3+
## New Features:
4+
5+
* Add some layers to Paddle
6+
* bilinear interpolation layer.
7+
* spatial pyramid-pool layer.
8+
* de-convolution layer.
9+
* maxout layer.
10+
* Support rectangle padding, stride, window and input for Pooling Operation.
11+
* Add —job=time in trainer, which can be used to print time info without compiler option -WITH_TIMER=ON.
12+
* Support Mac OS X Sierra by source code.
13+
* Expose cost_weight/nce_layer in `trainer_config_helpers`
14+
* Add FAQ, concepts, h-rnn docs.
15+
* Add Bidi-LSTM and DB-LSTM to quick start demo @alvations
16+
* Add usage track scripts.
17+
18+
## Improvements
19+
20+
* Add travis-ci for macos. Enable swig unittest in travis. Skip travis-ci when only docs are changed.
21+
* Add code coverage tools.
22+
* Refine convolution layer to speedup and reduce GPU memory.
23+
* Speed up PyDataProvider2
24+
* Add ubuntu deb package build scripts.
25+
* Make Paddle use git-flow branching model.
26+
* PServer support no parameter blocks.
27+
28+
## Bug Fixes
29+
30+
* add zlib link to py_paddle
31+
* add input sparse data check for sparse layer at runtime
32+
* Bug fix for sparse matrix multiplication
33+
* Fix floating-point overflow problem of tanh
34+
* Fix some nvcc compile options
35+
* Fix a bug in yield dictionary in DataProvider
36+
* Fix SRL hang when exit.
37+
38+
# Release v0.8.0beta.1
39+
New features:
40+
41+
* Mac OSX is supported by source code. #138
42+
* Both GPU and CPU versions of PaddlePaddle are supported.
43+
44+
* Support CUDA 8.0
45+
46+
* Enhance `PyDataProvider2`
47+
* Add dictionary yield format. `PyDataProvider2` can yield a dictionary with key is data_layer's name, value is features.
48+
* Add `min_pool_size` to control memory pool in provider.
49+
50+
* Add `deb` install package & docker image for no_avx machines.
51+
* Especially for cloud computing and virtual machines
52+
53+
* Automatically disable `avx` instructions in cmake when machine's CPU don't support `avx` instructions.
54+
55+
* Add Parallel NN api in trainer_config_helpers.
56+
57+
* Add `travis ci` for Github
58+
59+
Bug fixes:
60+
61+
* Several bugs in trainer_config_helpers. Also complete the unittest for trainer_config_helpers
62+
* Check if PaddlePaddle is installed when unittest.
63+
* Fix bugs in GTX series GPU
64+
* Fix bug in MultinomialSampler
65+
66+
Also more documentation was written since last release.
67+
68+
# Release v0.8.0beta.0
69+
70+
PaddlePaddle v0.8.0beta.0 release. The install package is not stable yet and it's a pre-release version.

0 commit comments

Comments
 (0)