- 21 12月, 2021 16 次提交
-
-
由 baoachun 提交于
* update seqconv_eltadd_relu_fuse_pass ut * update ut * update ut * update ut
-
由 baoachun 提交于
* update squared_mat_sub_fuse_pass ut * update ut * update ut
-
由 Chen Weihang 提交于
* rename cuda to gpu * revert CMake change * resolve conflit * rename other cuda to gpu * poish details
-
由 crystal 提交于
* relu forward opt * add gelu functor * optimize code
-
由 arlesniak 提交于
-
由 Yuang Liu 提交于
-
由 Guoxia Wang 提交于
-
由 baoachun 提交于
* add seqpool_cvm_concat_fuse_pass ut * rename ut name
-
由 chentianyu03 提交于
* fix when out_dtype is same with x.dtype and still transform type error * fix spell error
-
由 sneaxiy 提交于
* mean first version * fix scalar mean * add fp16 dtype for api
-
由 yeliang2258 提交于
* fix timeout bug * update
-
由 baoachun 提交于
* update repeated_fc_relu_fuse_pass ut * update ut
-
由 Haohongxiang 提交于
* update * fix bugs * modify code style * fix bugs of _get_global_group
-
由 heliqi 提交于
* add timeout * add timeout * PassAutoScan base_line use same config * try run base_line * fix dropout Mask of output attr error * fix dropout Mask of output attr error
-
由 石晓伟 提交于
* updates the check_file_diff_approvals for allocation refactor, test=develop * fix a bug, test=develop
-
由 Chen Weihang 提交于
* remove eigen and blas dir * fix declare error
-
- 20 12月, 2021 20 次提交
-
-
由 sneaxiy 提交于
-
由 baoachun 提交于
* add mkldnn conv_transpose_bias fuse pass ut * update conv_transpose_bias_mkldnn_fuse_pass ut * update conv_transpose_bias_mkldnn_fuse_pass ut * update conv_transpose_bias_mkldnn_fuse_pass ut * restrict conv2d data_format in conv_transpose_bias_mkldnn_fuse_pass * update ut timeout setting * update ut
-
由 chentianyu03 提交于
* add pten conj kernel * modify conj_kernel file path * add defined cuda macro to cuda/conj_kernel.h
-
由 baoachun 提交于
-
由 fwenguang 提交于
-
由 From00 提交于
-
由 sneaxiy 提交于
* support FP16 for more ops * add amp list tests * refine reduce_mean_grad * fix OP benchmark ci * fix fp16 reduce_mean * updat ut, but still have some problems * remove mean/reduce_mean fp16 kernel
-
由 Feng Xing 提交于
softmax_with_cross_entropy optimization with soft label. This PR includes optimization of "SoftmaxWithCrossEntropySoftLabel" : compute log_softmax and then compute loss. "CrossEntropySoftLabel" : compute loss with softmax as input. These optimization includes following technics: read data to buffer with vectorization compute max and sum in warp fixed loop size with macro Performance (computation time): softmax_with_cross_entropy_0 (forward) : -40.1% softmax_with_cross_entropy_0 (backward): -41%
-
由 石晓伟 提交于
-
由 Feiyu Chan 提交于
-
由 heliqi 提交于
* add matmul_scale matmul_v2_scale fuse pass * add scaletensor judge * modify var name * add timeout notest;test=coverag * fix error commit * fix use_mkldnn attr * fix use_mkldnn attr
-
由 Sylwester Fraczek 提交于
-
由 Sing_chan 提交于
* test if windows still need numpy <=1.19 * modify acoording to zhouwei's comment
-
由 kuizhiqing 提交于
-
由 0x45f 提交于
-
由 zhangbo9674 提交于
* add multi_tensor for momentum and clear_grads for optimizer * fix bug for dygraph * add unittest * refine comment * add param_group * refine regularizaiton logic * del clear_grads * add clear_grads * add dispensable check of None * refine clear_grad * fix build bug * refine code by comment * refine code * add multi tensor check * refine param_group update * add multi tensor for static mode * refine comments * delete useless comma for momentum * refine comment for momentum * refine code by commment
-
由 Yuang Liu 提交于
-
由 Feiyu Chan 提交于
-
由 YuanRisheng 提交于
* fix bugs when run reshape * fix ci bug
-
由 zyfncg 提交于
-
- 19 12月, 2021 1 次提交
-
-
由 Baibaifan 提交于
-
- 18 12月, 2021 3 次提交
-
-
由 Noel 提交于
-
由 Guoxia Wang 提交于
-
由 yeliang2258 提交于
* add test_conv_act_mkldnn_fuse_pass * update cmakelist * fix cmakelist * fix timeout * fix timeout * fix timeout * fix
-