30211 Commits (587d99ae443c684faa25d1fd261eb81d37cb32e4)

 

All Branches   

Search

Author	SHA1	Message	Date
Zhang Ting	7d95e598c1	support float16 for temporal_shift op (#31432)	4 years ago
YUNSHEN XIE	3a8ef10e09	fix modified_retry_method_only_win (#31404) ... * fix modified_retry_method_only_win * fix bug * fix retry bug on windows	4 years ago
Zhang Ting	dcce54ea76	improve performance of depthwise_conv2d (#31099) ... * improve performance of depthwise_conv2d * add unittest	4 years ago
wuhuanzhou	4d6d2db812	Windows system supports Ninja compilation (#31161)	4 years ago
liym27	0fff930667	Fix bug for set_value op when input dtype is not float32 (#31411)	4 years ago
Huihuang Zheng	c40b98e068	Fix comment (#31424) ... Fix wrong code comment	4 years ago
Huihuang Zheng	6bf02a1261	[Dy2stat] Fix Read-Only Attribute as while_loop Output (#31415) ... Fix Read-Only Attribute as while_loop Output: Usually, our convert_while_loop will be like: ``` [a, b, c] = paddle.jit.dy2static.convert_while_loop( condition_name, body_name, [a, b, c]) ``` where a, b, c are in loop_var_names. However, if loop_var_names contains property such as foo.x, we cannot assign the attribute as output of convert_while_loop because Python property is a kind of read-only attribute. To handle the case, we replace the attributes which are output of convert_while_loop with generated variables, then if we know the attribute is not read-only at runtime, we assign the attribute. The created statements are like: ``` [a, b, __attribute_variable_1] = paddle.jit.dy2static.convert_while_loop( condition_name, body_name, [a, b, foo.x]) if not isinstance(getattr(type(foo), x, None), property): foo.x = __attribute_variable_1 ```	4 years ago
jakpiase	5b4f8aac82	Added LSTM BF16 and fixed GRU BF16 (#31234)	4 years ago
Qi Li	7cdf6ea770	[ROCM] update fluid elementwise op for rocm (part10), test=develop (#31361) ... * [ROCM] update fluid elementwise op for rocm (part10), test=develop * update, test=develop * address review comments, test=develop	4 years ago
Qi Li	84639b6193	[ROCM] update fluid operators for rocm (part3), test=develop (#31213) ... * [ROCM] update fluid operators for rocm (part3), test=develop * fix clang format error, test=develop	4 years ago
Qi Li	3b9db17199	[ROCM] update fluid operators for rocm (part7), test=develop (#31307)	4 years ago
Qi Li	db50fb6766	[ROCM] fix softmax with loss and update python scripts, test=develop (#31373)	4 years ago
Pei Yang	32211fe9c4	TRT conv2d converter support SAME padding (#31379)	4 years ago
Qi Li	e312a1ff6e	[ROCM] update fluid operators for rocm (part9), test=develop (#31338)	4 years ago
Qi Li	6626c6a6ad	fix bert cu file compiler error, test=develop (#31389)	4 years ago
wuhuanzhou	c1bc223695	compile with VS2017, test=develop (#31388)	4 years ago
Zhou Wei	13e4280f82	[Custom OP]polish doc of custom OP (#31369)	4 years ago
Qi Li	946dbdae8c	[ROCM] update fluid operators for rocm (part6), test=develop (#31301)	4 years ago
wangna11BD	1cbccfa594	Add attrs `deformable_groups` for deformable_conv API (#31335) ... * add attrs deformable_groups	4 years ago
Shang Zhizhou	77c44e2f1b	change prelu plugin to tensorRT layer (#30210)	4 years ago
YUNSHEN XIE	353dd0cd98	Modified retry method on windows (#31363)	4 years ago
Qi Li	59940cb383	[ROCM] update fluid operators for rocm (part8), test=develop (#31309)	4 years ago
tangwei12	5d7a8b05f8	fix sycn training error (#31357) ... * fix sycn training error Change-Id: Ie2feebcf0b5b2984fd59cfcdde0c817840e203d2	4 years ago
Qi Li	ec72f5b235	fix ELU output for nan, test=develop (#31132)	4 years ago
Qi Li	65bcaeb004	[ROCM] update fluid operators for rocm (part5), test=develop (#31258) ... * [ROCM] update fluid operators for rocm (part5), test=develop * address review comments, test=develop * fix typo, test=develop	4 years ago
YUNSHEN XIE	2111d912d4	Decrease threshold for failed ut retry (#30903) ... * Decrease threshold for failed ut retry * retry Method upgrade * second method upgrade * fix error * Remove the comment lines * test for modified_retry_times * fix error * fix some error * fix error * fix error * remove test content * fix error * Reduce duplicate code * fix more than 10 ut failed bug * fix more than 10 ut failed bug on mac	4 years ago
Pei Yang	2e9e3fad15	add n-d input support for trt scale converter (#31316) ... * add n-d input support for trt scale converter * add flatten for ut * fix dims	4 years ago
Shang Zhizhou	6404c43814	support trt serialize when load model from memory (#31342) ... * support trt serialize when load model from memory * delete conv_bn_fuse_pass before tensorrt, with which trt serialize engine id is not stable * Revert "delete conv_bn_fuse_pass before tensorrt, with which trt serialize engine id is not stable" performance degradation, fix in the future This reverts commit fa6cd17e60b15df351efda379ddd00e9e9c1fea9. * add delete conv_bn * delete path when delete_cache_files	4 years ago
chentianyu03	a2c0b60401	remove wlist_temp in wlist.json (#31356)	4 years ago
Gradie	d79fdc3d62	lamb_op_xpu;test=kunlun (#31012) ... * lamb_op_xpu;test=kunlun * modify lamb_op_xpu.cc;test=kunlun * delete atol lamb_op_xpu; test=kunlun * update xpu.cmake;test=kunlun * test_error 1e-5,lamb_op_xpu;test=kunlun * error1e-5,lamb_op_xpu,test=kunlun * delete atol lamb_xpu;test=kunlun * modify atol,lamb_op_xpy;test=kunlun * lamb_op_xpu;test=kunlun * lamb_op_xpu;test=kunlun * lamb_op_xpu, XPUOptest;test=kunlun * lamb_op_xpu;test=kunlun * lamb_op_xpu;test=kunlun * lamb_op_xpu;test=kunlun * lamb_op_xpu;test=kunlun * lamb_op_xpu;test=kunlun * lamb_op_xpu;test=kunlun * lamb_op_xpu;test=kunlun * lamb_op_xpu;test=kunlun * lamb_op_xpu;test=kunlun * lamb_op_xpu;test=kunlun * lamb_op_xpu;test=kunlun * lamb_op_xpu,modify xpu_cmake; test=kunlun * lamb_op_xpu;test=kunlun * lamb_op_xpu,modify xpucmake;test=kunlun	4 years ago
danleifeng	d1075df2e8	topo and memory performance for heterps (#30440) ... * topo and memory performance for heterps; test=develop * add trainwithprofiler in heter trainier; test=develop	4 years ago
Qi Li	72d99c5dcd	[ROCM] update fluid operators for rocm (part4), test=develop (#31225)	4 years ago
cucuzg	91635de390	opt matmul and matmul_v2 on kunlun, *test=kunlun (#31326) ... * add clip_by_norm on kunlun, *test=kunlun * opt matmul and matmul_v2 on kunlun, *test=kunlun	4 years ago
Wilber	e20234094c	Fix xpu compile and cipher symbol problem. (#31271)	4 years ago
wuhuanzhou	30858d8974	fix compilation errors for missing brpc header files, test=develop (#31325)	4 years ago
石晓伟	625482f752	inference modification for custom operator, test=develop (#31312)	4 years ago
wuhuanzhou	a13f1d6930	optimize unity build (#31119) ... * optimize unity build, test=develop * fix compilation error on Windows, test=develop * fix compilation error, test=develop * fix code style error, test=develop	4 years ago
jiangcheng	8f4ac6b525	optimize topk op through limit SortTopK kernel entrance, test=develop (#30403)	4 years ago
alncat	bfb8a64234	updated conv bn fuse pass to make it compatible with latest batch_norm op (#31272)	4 years ago
Bin Lu	a37658daff	Update transforms.py (#31252) ... update RandomHorizontal Flip random description and examples	4 years ago
Chen Weihang	5610c1717e	fix dtype unmatched (#31305)	4 years ago
Chen Long	f8bdb90917	fix readme test=document_fix (#31308)	4 years ago
Qi Li	9b016c7cb7	[ROCM] update fluid operators for rocm (part2), test=develop (#31211)	4 years ago
niuliling123	2fd999d979	Optimized the adaptive_avg_pool2d op when output_size == 1 (#31197) ... * Optimized the adaptive_avg_pool2d op when output_size == 1	4 years ago
Zhou Wei	aebf223478	fix test_check_abi (#31288)	4 years ago
石晓伟	1da3280660	inference modification for custom operator, test=develop (#31283)	4 years ago
Zhou Wei	cc89120a2c	[Custom OP]add MSVC compile check on Windows (#31265)	4 years ago
Zhou Wei	af9066e89c	[Custom OP]add PD_THROW and PD_CHECK for User Error message (#31253) ... * [Custom OP]add PD_THROW and PD_CHECK for User error message * PD_THROW and PD_CHECK, fix comment * fix Windows error message * fix Windows error message * fix CI	4 years ago
石晓伟	8c94d8cb4c	[Custom OP] change the user header file format, test=develop (#31274)	4 years ago
Jiabin Yang	038ce70d69	[Custom OP] Support stream set on Custom Op (#31257)	4 years ago