15 Commits (5a3b06bef7d02366c5fc1055c5d5b6fe1a241c75)

Author	SHA1	Message	Date
QI JUN	61ec0b9516	Refine device context (#6433) ... There are mainly following fixes: - take `DeviceContext` as the template parameter of math functors and OpKernel instead of `Place` - remove `eigen_device` interface in base class `DeviceContext` - remove `GetEigenDevice` interface in `ExecutionContext` and base class `DeviceContext` - remove unused `platform::EigenDeviceConverter` - rename `REGISTER_OP_GPU_KERNEL` to `REGISTER_OP_CUDA_KERNEL` - rename `USE_GPU_ONLY_OP` to `USE_CUDA_ONLY_OP`	7 years ago
chengduoZH	5238c9fb0b	input type should be different	7 years ago
chengduoZH	2160497786	fix pooling functor parameter order	7 years ago
chengduoZH	6c6474cbd8	follow coments	7 years ago
chengduoZH	fcfce48421	follow coments	7 years ago
chengduoZH	bb33c2b3a5	fix kernel func	8 years ago
chengduoZH	e1e3859e88	remove custom attr checker and fix code format	8 years ago
chengduoZH	dfc8d3c1c1	Fix (According to the review)	8 years ago
chengduoZH	905a462d3c	Merge branch 'fix_maxpool_backward_functor_temp' into Add_pool_op_temp	8 years ago
chengduoZH	b72854389e	Fix (According to the review)	8 years ago
chengduoZH	9b1431b8d9	fix maxpool backward functor and add unit test	8 years ago
chengduoZH	0417e4e4bf	fix framework::LoDTensor => Tensor	8 years ago
chengduoZH	84a2512b90	fix parameter name and function define	8 years ago
chengduoZH	3416f5e0f8	fix function define	8 years ago
chengduoZH	9e7c0b5ef7	Add pooling2d(max, ave) and pooling3d(max, ave) functor	8 years ago