2831 Commits (c9ea317b367b61c3abc6496e76354fb297b4746c)

Author	SHA1	Message	Date
tangwei12	8f0b3c0516	the integrated communicator (#19849) ... * add a base class for the Communicator * add AsyncCommunicator Impl for async distributed training	6 years ago
zhaoyuchen2018	5a2ecdea40	Fix name_scope test case bug (#20034) ... test=develop Signed-off-by: zhaoyuchen <zhaoyuchen01@baidu.com>	6 years ago
gongweibao	afc40a59eb	Add `RUN_SERIAL` attribute to `exclusive` test. (#20026)	6 years ago
wangchaochaohu	3409db950c	fix reduce bug test=develop (#19971)	6 years ago
liuwei1031	bda7eab77d	improve the error message when handling ndarray with unsupported dtype (#19949) ... * impove error message when passing ndarray with object dtype * imporve message format * change assert to raise TypeError * remind user how to locate the irregular data instead of printing * add unittest for input array type check	6 years ago
joanna.wozna.intel	1d32897c5c	Fix test pool2d int8 mkldnn (#19976) ... * Fix conv2d+dequantize squash for residual fusion test=develop * Correct int8 input test=develop * Add if exclude or include padding in pool2d mkldnn test=develop	6 years ago
Aurelius84	f58c8db668	Require x.dims=label.dims in huber_loss (#20017) ... * x.dims == y.dims test=develop * refine comment	6 years ago
Yang Zhang	cde73a7bbf	Expose `mutable_data` as python binding (#19932) ... * Expose `mutable_data` as python binding test=develop * Add test for device pointer binding test=develop * Make test compatible with python 2	6 years ago
Aurelius84	137e6336ef	Remove constraint that last dimension is forced to be 1 in rank_loss (#19997) ... * fix input shape check test=develop * move PADDLE_ENFORCE test=develop	6 years ago
chengduo	101a2b610a	Add dtype for coalesce_tensor_op (#20016) ... Add dtype for coalesce_tensor_op	6 years ago
Huihuang Zheng	88af4ab650	Add new data layer (#19916) ... The new "fluid.data" changes old "fluid.layers.data": 1. Add shape and dtype check. 2. Remove "append_batch_size" parameter. We won't offer this in the new data layer because other deep learning platforms don't have this kind of data layer pre-processing. It may confuse users. 3. Remove "stop gradient" parameter because the data layer doesn't do back-propagation TODO： Now data layer feeded by executor is checked, will we want to check the feed data of readers in the future?	6 years ago
Zeng Jinle	1b7de89455	fix math_op_path.py when integers, test=develop (#20008)	6 years ago
qingqing01	bb271b6d44	Remove unit testing for large shape in test_affine_channel_op (#19993)	6 years ago
zhongpu	b1bb23841e	add kernel for fill_op, test=develop (#19719) ... * add kernel for fill_op, test=develop * modify PADDLE_ENFORCE to PADDLE_ENFORCE_EQ, test=develop * add op test for fill_op, test=develop * REGISTER COP CUDA KERNEL, test=develop * update test_fill_op.py, test=develop * change FillConstantOpVarTypeInference to FillOpVarTypeInference, test=develop * fix op test, test=develop * add head file, test=develop	6 years ago
wangchaochaohu	382d099dcb	add support tensor and tensorlist for strided_slice OP (#19929) ... * add support tensor and tensorlist for strided_slice OP test=develop * fix the commnet test=develop * fix test=develop * fix the bug test=develop * delete log test=develop * fix API.spec test=develop * fix test=develop	6 years ago
lvmengsi	619a241bd0	Fix OpTest of bn (#19062) ... * fix bn	6 years ago
ShenLiang	5920d69df5	Avoid treating broadcast as initialization operation (#19857) ... * treat broadcast as non-initial, test=develop * rename the class name * rename the class name, test=develop	6 years ago
Bob Zhu	c670058a8d	add support of matmul with multiple head even different width and height (#19708) ... * add support of matmul with multiple head even different width and height Original matmul with multiple head supports only the mat_a.width == mat_b.height, in that case, mat_b will be horizontally split. In this patch, we extend the support when mat_a.width != mat_b.height but mat_a.width/head_number == mat_b.height, in this case, mab_b will be vertically split. One example is A is [3, 8], B is [2, 16], head_number is 4. In this case, A will be split as [3, 2], B will be (vertically) split as [2, 4]. The final result will be 4 matrix of 4 matrix of [3,4], i.e. [3, 16] test=develop * add support of matmul with multiple head even different width and height Original matmul with multiple head supports only the mat_a.width == mat_b.height, in that case, mat_b will be horizontally split. In this patch, we extend the support when mat_a.width != mat_b.height but mat_a.width/head_number == mat_b.height, in this case, mab_b will be vertically split. One example is A is [3, 8], B is [2, 16], head_number is 4. In this case, A will be split as [3, 2], B will be (vertically) split as [2, 4]. The final result will be 4 matrix of 4 matrix of [3,4], i.e. [3, 16] test=develop * refactor the code of matmul with multiple head even different width and height test=develop	6 years ago
Liufang Sang	6884dc800a	refine ctc align op with padding (#19926) ... * refine ctc align op with padding * refine api sample code	6 years ago
Tao Luo	65a02fc114	add input type and dtype check for softmax_op (#19975) ... * add input type and dtype check for softmax_op test=develop * refine error message test=develop	6 years ago
Aurelius84	99a9615a4b	Removing length dims constraints of seq_pad and seq_unpad (#19497) ... * Removing last dims constraints of seq_pad and seq_unpad test=develop * fix test_layer api code test=develop * fix sequence_pad_op.cc conflict test=develop * remove test_analyzer_mm_dnn test=develop * fix vectorize bug test=develop * fix vectorize<int> test=develop	6 years ago
jhjiangcs	766bd529d1	add optimizer:dpsgd,test=develop (#19915)	6 years ago
Yang Zhang	ebff68fa74	Add float16 support to `sync_batch_norm_op` (#19681) ... * Add float16 support to `sync_batch_norm_op` test=develop * Add test for sync_bn with FP16 input test=develop	6 years ago
Aurelius84	039b9710d5	Remove constraint that last dimension is forced to be 1 by adding lookup_table_v2 (#19735) ... * Remove constraint that last dimension is forced to be 1 by add lookup_table_v2 test=develop * modify into PADDLE_ENFORCE_CUDA_SUCCESS test=develop * Revert "modify into PADDLE_ENFORCE_CUDA_SUCCESS test=develop" This reverts commit 8a960bfc61e51aa27c3c529df8fb90b93ebd19f9. * move api into fluid.embedding test=develop * fix example code test=develop * move one_hot into fluid.one_hot * modify api.spec test=develop * fix loss shape test=develop	6 years ago
xujiaqi01	cedc04775c	support change shuffle and train thread num (#19841) ... * support change shuffle thread num * support change train thread num * fix receive shuffle data of each channel * data norm stop gradient * add check thread_tensor type and root_tensor type when merge metric * remove sleep in shuffle, add config * add config of pslib client to client communication * fix xbox str * add data norm op testcase * add flush in trainer finalize	6 years ago
Kaipeng Deng	14625ffe9e	add elementwise mod support float/double. test=develop (#19570)	6 years ago
Ghost Under Moon	790d5226b5	give warnings when save a model without any parameters (#19931) ... * give warnings when save a model without any parameters test=develop * delete one line comment test=develop	6 years ago
Zeng Jinle	f254b477d1	Add py_reader combination unittest (#19923) ... * add py_reader combination unittest,test=develop * follow huihuang's comments, test=develop	6 years ago
Leo Chen	57606205f5	Make OpTest check grad inplace even if forward has no inplace (#19847) ... * make OpTest check grad inplace even if forward has no inplace, test=develop * do not run PE when enable_inplace is False, test=develop * add conv3d cuda kernel for float16 type, test=develop * refactor OpTest for inplace, test=develop * add comments, test=develop	6 years ago
Zhang Ting	cb8f3c03a7	resize Ops support data_layout:channel_last, test=develop, test=document_preview (#19914)	6 years ago
mapingshuo	9901f69677	Forward recompute3 (#19913) ... * add recompute based checkpoints methods for large batch training test=develop * add append_backward_with_forward_recomputation test=develop * refine optimizer test=develop * update backward and optimizer test=develop * make Variable usable test=develop * add recompute code * refine optimizer test=develop * refine addup _append_backward_ops_with_checkpoints_ 1) for recompute part, just cache the grad_op_desc without appending to block 2) before appending grad_op_desc to backward part, addup_repetitive_vars, remove unused branch test=develop * make method private * add recompute strategy into DistributedStrategy test=develop * checkpoint version3 test=develop * remove some print information test=develop * remove unused sumop test=develop * try to fix recompute with graph building modules * add input names to vars should be held * add memory debug tool * backup backward * Fix bugs * add backward desc for op not in any segments * add exception info for sub_block test=develop * modify code style test=develop * modify code style test=develop * remove print functions test=develop * add API spec test=develop test=document_preview * make Recompute a child class of Optimizer test=develop test=document_preview * add API spec test=develop test=document_preview * modify API spec test=develop test=document_preview * add document for Recompute test=develop test=document_preview * change API doc of Rcompute test=develop test=document_preview * code cleaning test=develop test=document_preview * modify API spec * fix bugs when segments hold no element * add testcase for Recompute Optimizer test=develop test=document_preview * add test for apply_gradient, and code cleaning test=develop test=document_preview * add test case for load function * enable CI test=develop test=document * add test case test=develop test=document_preview * add sample code for 4 function of recompute optimizer test=develop test=document_preview	6 years ago
Ghost Under Moon	4836ee68fb	warning when user save a inference model which contains auc op test=develop (#19838)	6 years ago
wopeizl	e606b1754e	optimize the error information when the input for while op has a wron… (#19872) ... * optimize the error information when the input for while op has a wrong shape test=develop	6 years ago
ruri	d31c92a2cd	add mse_loss (#19759) ... * add mse_loss op	6 years ago
Tao Luo	a4919d3688	move tree_conv to fluid.contrib.layers (#19918) ... * move tree_conv to fluid.contrib.layers test=develop * update API.spec for tree_conv test=develop * update tree_conv api to increase unit coverage test=develop	6 years ago
Zeng Jinle	0436efd6a3	Unify DataLoader APIs (#19305) ... * unify DataLoader APIs, test=develop * integrate iterable CPU Dataset, test=develop add GPU dataset supporting, test=develop * add unittests for dataset, test=develop * add more docs to dataloader apis, test=develop, test=document_preview * refine doc, test=develop * refine doc again, test=develop * increase coverage, test=develop	6 years ago
tangwei12	278dd00322	paddle cloud role maker fix (#19646) ... * optimize cloud rolemaker, test=develop	6 years ago
lvmengsi	4155e62559	add instance norm (#19500) ... * add instance norm op	6 years ago
Adam	cb65439da8	Add support for other axes in MKLDNN softmax op (#19907) ... * Initial, functional commit * Clean commit related files test=develop	6 years ago
Jiabin Yang	454254115e	Feature/auto prune in dygraph (#19757) ... * refactor dygraph,test=develop * fix failed unittest,test=develop * polish code,test=develop * check windows ci error,test=develop try to fix windows ci error by np.allclose,test=develop * polish vlog and profiler, test=develop * try to fix preceding ops order,test=develop * test transformer in windows ci, test=develop * use python c-api to speed up tracer.trace,test=develop * test=develop, fix docker with paddle nccl problem * test=develop, add ut for debug string and gradient_accumulator * test=develop, add tests for layer/gradient_accumulator/prepared_op * test=develop, fix complie error for test_prepared_op * test=develop, add more ut for dygraph * test=develop, create API.spec for dygraph api change * test=develop, refoctor name to make it easier to understand * test=develop, refoctor name to make it easier to understand * test=develop, fix multi-gpu failed problem , add Tracer tests, change PADDLEENFORCE to PADDLEENFORCE_EQ * test=develop, fix ut failed on parallel se-resnext * test=develop, change one more PADDLE_ENFORCE * support auto prune in dygraph mode * test=develop, support auto prune * test=develop, merge develop conflict * test=develop, fix test_layer and test_tracer ut * test=develop, fix bug which may cause stop_gradient disabled with a list of backward inputs	6 years ago
Aurelius84	418a0967f3	move match_matrix var_conv2d et.al api into fluid.contrib test=develop (#19859)	6 years ago
Zeng Jinle	cee0079a4d	fix readers bug, test=develop (#19868)	6 years ago
Aurelius84	fcf53e55ff	support 2-level lod of input in sequence_pool (#19839) ... * support 2-level lod of input in sequence_pool test=develop * fix lod level bug in .cu test=develop	6 years ago
Zhang Ting	93364b45c1	group_norm support data_layout:NHWC, test=develop, test=document_preview (#19614) ... 1. group_norm support data_layout=NHWC 2. modified doc of group_norm	6 years ago
Zhang Ting	439d95e157	modified interpolate op to support tensor attribute, test=develop, test=document_preview (#19287) ... modified interpolate_op to support tensor attribute 1. the parameter out_shape of image_resize、resize_nearest/bilinear/trilinear can be a list or a 1-D tensor variable. If a list, each element can be an integer or a tensor variable with shape: [1]. 2. the parameter scale of above Ops can be a 1-D tensor variable. modified document of image_resize, resize_nearest, resize_bilinear, resize_trilinear and add some code example.	6 years ago
Zhang Ting	b38889413d	add crop_tensor_op, test=develop, test=document_preview (#19314) ... add crop_tensor op. The main difference with crop is : 1. If the argument shape is a list, each element is an integer or a tensor variable with shape: [1]. This way is suitable for the case that the shape may be changed each iteration. 2. If the argument shape is a variable. Its rank must be 1. In crop op, the rank of shape must be the same as x offsets can be a list, in which each element is an integer or a tensor variavle with shape: [1].	6 years ago
flame	fe18cfdb4f	hide with inference optim API (#17355)	6 years ago
Aurelius84	b125e327aa	Remove constraint that last dimension is forced to be 1 in cross_entropy (#19606) ... * Remove constraint that last dimension is forced to be 1 in cross_entropy test=develop * modify labels last dims test=develop	6 years ago
wopeizl	a7c440d303	add precise roi pooling op test=develop (#18960) ... * add precise roi pooling op test=develop * test=develop * test=develop * test=develop * test=develop * test=develop * test=develop * test=develop * test=develop * test=develop * test=develop * test=develop * test=develop * test=develop * detail the description test=develop * test=develop * elaborate the doc for return type test=develop * test=develop	6 years ago
Yiqun Liu	3cd985a669	Add a pass to fuse fc+elementwise_add+layernorm (#19776) ... * Add fc_elementwise_layernorm_fuse pass and unittest. * Add fused_fc_elementwise_layernorm op and its GPU kernel. test=develop * Apply fc_elementwise_layernorm_fuse_pass to GPU inference. * Add the setting of attrs in the definition of binary_op. test=develop * Add comment. * Implement the unittest. test=develop * Change the unittest name of layer_norm. test=develop	6 years ago