70 Commits (e05f4ff26700dd34aa6d3c6da7061c62c5fa39c9)

Author	SHA1	Message	Date
Yu Yang	e05f4ff267	Fix SRL hang when exit. (#291) ... * Fix SRL hang when exit. * Error occurred when enable Async Load in TestDataProvider. * It because DataProvider is calling getNextBatchInternal in one thread, and destructing DataProvider in other thread. * Add wait routine in DataProvider destructing. * Also fix another bug, when destructing TestDataProvider and do not read any test data. Fix #286 * Follow comments, Use mutex is cool!	8 years ago
emailweixu	6c3a678c9a	Add elementwise math operations (#343) ... * Add elementwise math operations This allows use to use expressions like: y=log(1+exp(x)) Also added unittests for ActivationFunction * Enforce keyword arguments for non-positional arguments * Add LogActivation to doc	8 years ago
hedaoyuan	1de75c039f	report error when use parallel_nn to train recurrent_nn model (#335)	8 years ago
qingqing01	5acf136615	Bug fix in CudnnConvLayer, which will lead to destruction error. (#317)	8 years ago
qingqing01	45c81a414f	Add job=time in trainer, refine cudnn_conv to reduce gpu memory and speed up training. (#218) ... * Add benchmark for PaddlePaddle, tensorflow and caffe * ConvProjection to reduce memory for goolenet * Add unit test for ConvProjection. 1. unit test in test_LayerGrad. 2. compare the ConvPorjection and CudnnConvLayer, also compare the concat_layer+img_conv_layer and concat_layer_conv_projection. * Reduce cudnn_conv memory and add benchmark document. 1. Use TmpMatrix as the workspace in cudnn_conv to reduce gpu memory. It reduce lots of memory. 2. Add benchmark document. 3. fix smallnet_mnist_cifar.py in paddle. * Add job=time and refine cudnn_conv to reduce gpu memroy and speed up * Refine cudnn_conv and shared biases operation in concat_layer and mixed_layer. * follow comments * follow comments * Use unique_ptr to prevent memory leaks in CudnnConvLayer.	8 years ago
gangliao	9c5c38fa2a	FIx check type unmatch in MaxOutLayer (#242) ... Compiled failed on gcc 4.6	8 years ago
luotao1	3dd8c9bea4	add maxout layer, including interface and unittest (#229) ... * add maxout layer, including interface and unittest * follow maxout comments * auto setting channels * fix unittest bug in test_RecurrentGradientMachine	8 years ago
luotao1	e1f57bfd66	add base class for seqlastin/max/average layer (#187)	8 years ago
luotao1	6bef8390b9	fix bug in some different python environment (#220)	8 years ago
Yu Yang	58f896c3f4	Speed up PyDP2, support numpy.float array (#207)	8 years ago
emailweixu	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	8 years ago
Yu Yang	2f82d72ede	Fix bug in yield dictionary in DataProvider. (#197) ... * Fix bug in yield dictionary in DataProvider. * Also make virtualenv work in Paddle.	8 years ago
luotao1	91df606280	remove some copyfrom in AgentLayer and ExpandLayer, fix warning in seq2seq config (#183)	8 years ago
luotao1	cebdb66768	hierarchical rnn document, add new config example (#106) ... * hierarchical rnn document, add new config example * update inputs_type of label * add check for unsupported config * refine hierarchical document * refine doc title * update docs, fix paddle to PaddlePaddle * follow comments	8 years ago
Yu Yang	c2d418dbfd	Merge internal changes (#198) ... * fix DataProvider create function args bug Change-Id: I9e3a1c535c805bf30204a14aea8d5143ff534784 * remove PserverForPython.h which is not used Change-Id: I2b27f1f3c11a42766a92fc689f0f5f1f73ee1d70 * add internal document script Change-Id: Ia0fec79456caea0b271f9903cc13e8a3d32e0774	8 years ago
luotao1	43f7d7b768	add interface and unittest for nce layer (#180) ... * add interface and unittest for nce layer * follow comments	8 years ago
qingqing01	191fafe355	support rectangle padding, stride, window and input for PoolProjection (#115) ... * support rectangle padding, stride, window and input for PoolProjection * Follow comments. 1. Remove start 2. refine img_pool_a/b.conf for test_NetworkCompare 3. Split unit test * Modify the test in img_layers.py	8 years ago
Zrachel	1c2ebe467b	add interface and test of RecurrentGradientMachine (#156) ... * add interface and unittest of RecurrentGradientMachine for the function of multiple Subsequence inlinks with unequal token length	8 years ago
hedaoyuan	b52039bd11	some bug fix for sparse matrix (#133) ... * some bug fix for sparse matrix * a minor bug fix	8 years ago
gangliao	2920b6bc0d	Merge pull request #138 from gangliao/master ... Add Mac OS X port	8 years ago
liaogang	8ddc5faac1	Update Mac OS X port ... * follow comments to fix bugs	8 years ago
emailweixu	6decbdf4f5	Fix potential dead lock in PyDataProvider2 (#140) ... This bug occasionally causes dead lock in test_RecurrentGradientMachine In general, conditional_variable::notify should be used together with mutex for changing condition.	8 years ago
liaogang	0072ef50bf	Fix compile check type failed in linux	8 years ago
liaogang	23e47bb600	Merge remote-tracking branch 'upstream/master'	8 years ago
liaogang	a8df411192	Replace random_shuffle using shuffle. ... * reduce trainer count for unit test on MAC OSX	8 years ago
emailweixu	b15a4783cb	Correctly handling multiple inputs and integer inputs for recurrent_g… (#114) ... * Correctly handling multiple inputs and integer inputs for recurrent_group * Fix ScatterAgentLayer for generation * Revert sequence_(nest)_rnn.conf	8 years ago
liuyuan	2289c141c2	Refine comment for CRF related headers. (#117)	8 years ago
liaogang	5bca34ed24	Merge remote-tracking branch 'upstream/master'	8 years ago
qingqing01	c1c07bbef6	Update CudnnConvLayer.cpp	8 years ago
dangqingqing	95da095d10	fix cudnn conv bug which occurs in image classfication demo in GTX GPU	8 years ago
Yu Yang	7eb29f264b	Try to fix MultinomialSampler (#102) ... * Also refine unittest to multiple iteration to prevent luckily random number.	8 years ago
liaogang	b8d26ff4d7	fix float except bugs	8 years ago
liaogang	7ff8e76229	Shrink batch size on unit test for Mac OS X	9 years ago
liaogang	9795e01639	Merge remote-tracking branch 'upstream/master'	9 years ago
Yu Yang	2c5a6ac095	Optional fields to shrink generated proto size (#93) ... * remove unnecessary field set in ParameterConfig, Evaluators, etc	9 years ago
Yu Yang	90b9cba7fe	Add min_pool_size, Add default value of should_shuffle (#70) ... * min_pool_size would be infinite by default. * add unittest for min_pool_size * Fix bug in can_over_batch_size * add unittest for can_over_batch_size * Add DEFINE_PROVIDER_EX * Add default value of should_shuffle * When training, the default value of should_shuffle is True. * When testing, the default value of should_shuffle is False. * User a set a provider should_shuffle or not by pass it to `@provider` * should_shuffle can handle a list of value, not just boolean * Add input order mapping by using name * Add unittest * Add check to check input format. * Default is close for speed reason. * User could stop train when check error, or continue train without this train sample. * use deque instead of vector in generators pool, make erase generator faster. * Add chinese/english documentation * Make should shuffle = false in unittest * Add python files to depends.	9 years ago
liaogang	a6b58cf013	Merge remote-tracking branch 'upstream/master'	9 years ago
liaogang	92ca98d5f5	Merge remote-tracking branch 'upstream/master'	9 years ago
xuwei06	48eb5ff01f	Fix signed/unsigned comparison for gcc 4.9	9 years ago
xuwei06	05a97ab50e	Further fix the memory for Hierarchical RNN ... Sequences should be sorted according to the number of subsequences they have.	9 years ago
qingqing01	df28da76c5	try to fix bug for CTCErrorEvaluator.cpp when batch_size > 1 (#82) ... * try to fix bug for ctc_error_evaluator	9 years ago
Yu Yang	c7762da330	Merge branch 'master' into merge_icode	9 years ago
xuwei06	9a9de9240d	Correctly handle memory in RecurrentGradientMachine for hirarchical RNN ... Change-Id: I8e0a8ea6fc2760652d9c76440a539c90860062d3	9 years ago
zhangruiqing01	699d5f2638	modify RecurrentGradientMachine to support unequal length inputs ... * modify RecurrentGradientMachine to support hasSubSeq sequence inlinks with the same number of sentence but different number of tokens for each sentence Change-Id: Ic71f00a4bb346b4fa93e650dfb4b1a0d8d2338b0	9 years ago
Haonan	0f91ea7ebb	use HPPL_STREAM_DEFAULT for layer computation ... Change-Id: Id66da7b7f5bf9ec80cc19b347e4fb822a5a6f197	9 years ago
Haonan	688eeefab3	fixed issues with synchronizing streams when copy from gpu to cpu ... * by default, synchronize default_stream after resizeAndCopyFrom * add sync in some places after resizeAndCopyFrom using other streams	9 years ago
xuwei06	7481429f99	Change cos_sim to use CosSimLayer layer when size=1 and rename convex_comb_layer to linear_comb_layer ... Also add an unittest for layers.py, currently for a few things. More need to be added later.	9 years ago
qingqing01	7dbc092c87	fix cudnn version number for batch norm. (#71) ... * fix CUDNN_VERSION for backward of CudnnBatchNormLayer * fix cudnn version number for BatchNorm	9 years ago
Yu Yang	d9870c354f	Merge branch 'master' into mac_port	9 years ago
liaogang	eb5cf64b5e	reduce data_layer size of unit test to avoid cuda out of memory on MAC OS	9 years ago