24509f4af9
Fix the grammar in copyright. ( #8403 )
7 years ago
1ed5ae7a14
Fix comparing between signed and unsigned values ( #5328 )
7 years ago
cdd1da3424
Hack auc for dense vector labels ( #5274 )
...
* refine evaluator op types
* update
* follow comments
* update
* fix v2 mnist case
* fix v2 mnist case
* update
* update
* hack auc evaluator for dense vec
* follow comments
7 years ago
a523bea8e5
fix getType.
8 years ago
0b478e991c
follow comments.
8 years ago
faa4da4835
fix ctc edit distance in v2 API.
8 years ago
5f924d5d53
Follow comments.
8 years ago
597a58c3ef
Add DetectionMAPEvaluator.
8 years ago
f3eb9cb36a
Override getValueImpl and revise document
8 years ago
8411a73e87
overload several virtual functions to make ChunkEvaluator output multiple metrics
8 years ago
9087e3a9bb
Follow comments to use struct get return value.
8 years ago
e67aac1b02
Merge branch 'develop' of github.com:baidu/Paddle into feature/EvaluatorValue
8 years ago
2bb7fd8177
Remove padde namespace in some code.
8 years ago
b6ac64a36b
Remove unnecessary error default value.
8 years ago
b1ab8b56cf
Use plain C++ 03 to implement getStatsInfo.
8 years ago
8fded24c75
implement top k classification error in class matrix
8 years ago
d256512832
Rewrite code according to reviewer comments
8 years ago
e768721cd9
fix calculating totalScore2_ bug
8 years ago
413cbb8432
fix a variable definition bug
8 years ago
043859b5db
clean up code
8 years ago
5250b9b213
fix classificationError input
8 years ago
e00f06afa4
Add top-k error
8 years ago
bb751b7f2c
Add documentation for codes.
...
* also remove unused getValueStr
8 years ago
9d45aa4fb2
Merge branch 'develop' into feature/EvaluatorValue
8 years ago
ab8bb9738a
Merge pull request #1373 from reyoung/feature/remove_grad_printer_seq_info
...
Remove sequence information in GradientPrinter.
Fix #1376
8 years ago
be3adec35d
Merge pull request #1372 from reyoung/feature/unify_print_layer_evaluator
...
Unify print layer, ValuePrinter evaluator.
Fix #1378
8 years ago
948218dab9
Unify PrintLogic in PrintLayer/ValuePrinter.
8 years ago
1100366803
Add unittests & fix some bugs.
8 years ago
6deb33d9dc
Complete combined evaluator
8 years ago
9a48442536
Merge branch 'feature/remove_grad_printer_seq_info' into feature/EvaluatorValue
8 years ago
5fd49f6665
Remove sequence information in GradientPrinter.
...
* it is not necessary, because user could print sequence info
by ValuePrinter.
8 years ago
5b4e7d5cc7
complete value printer
8 years ago
5ecc1a2192
Merge branch 'feature/unify_print_layer_evaluator' into feature/EvaluatorValue
8 years ago
04eaf75c5e
Add getValue to some evaluators.
8 years ago
2c07dd50ab
Unify print layer, ValuePrinter evaluator.
8 years ago
39feacb02a
Merge branch 'feature/UseInitFuncInsteadCreate' into feature/EvaluatorValue
8 years ago
d2c58907eb
Unify Evaluator::create method
...
* Simplify Evaluator::create logic. Always use ClassRegister to
create Evaluator.
8 years ago
d5cfa6fcce
Stash
8 years ago
904936912e
Merge pull request #934 from tianbingsz/paddle_function_mat
...
Matrix API refactor
8 years ago
6e405a10c5
fix style issues
8 years ago
8fe3a3aa73
Add excluded_chunk_types to ChunkEvaluator
...
The chunks of types in excluded_chunk_types will not be counted in
ChunkEvaluator. This is useful for tasks such as SRL, in which chunks of
type V (verb) will not be taken into account in evaluation.
8 years ago
706c572424
Matrix API refactor, when passing parameters, convert shared_ptr (MatrixPtr) to
...
reference or raw matrix (Matrix & or Matrix *)
contextProjectionForward
contextProjectionBackward
contextProjectionBackwardData
contextProjectionBackwardWeight
classificationError
The mul functions would be updated later.
8 years ago
3d0e73bd32
Remove custom glog-like and gflags-like macros
8 years ago
068bfbb817
All file pass pre-commit hook
8 years ago
e9549cbb78
Change "Baidu, Inc" into "PaddlePaddle Authors"
8 years ago
80c68d38ff
clang format .cc .h .cpp .c and .hpp file
8 years ago
20600e7036
Fix several cpp issues
...
* Different Type compare.
* ostream << should pass a const object.
* remove always true checks.
8 years ago
28bc05b126
Fix sparse training for trainer_count=1 ( #204 )
...
* Fix sparse training for trainer_count=1
For trainer_count=1, the gradient machine is NeuralNetwork, which does not create parameter buf for PARAMETER_GRADIENT for sparse update in Parameter::enableType. But gradient parameter buf is still used in SgdThreadUpdater.
* Minor update to comment
9 years ago
df28da76c5
try to fix bug for CTCErrorEvaluator.cpp when batch_size > 1 ( #82 )
...
* try to fix bug for ctc_error_evaluator
9 years ago
0f91ea7ebb
use HPPL_STREAM_DEFAULT for layer computation
...
Change-Id: Id66da7b7f5bf9ec80cc19b347e4fb822a5a6f197
9 years ago
qingqing01
Yi Wang
武毅
caoying03
yangyaming
yangyaming
Yu Yang
Liang Zhao
tianbingsz
Peng Li
xutianbing
liaogang
Yi Wang
Luo Tao
emailweixu
Haonan