backward data (#7)

* enabled atomic add in tensor copy
* added gridwise GEMM
* added backward data conv using GEMM + atomic
* added backward data conv using GEMM, no atomic


[ROCm/composable_kernel commit: 8f5f64960e]
This commit is contained in:
Chao Liu
2019-12-03 01:16:12 -06:00
committed by GitHub
parent 809e8239e1
commit 4414e495ed
51 changed files with 3563 additions and 570 deletions

3
script/docker-cuda.sh Executable file
View File

@@ -0,0 +1,3 @@
WORKSPACE=$1
echo "workspace: " $WORKSPACE
sudo docker run -it -v $WORKSPACE:/root/workspace --group-add sudo --runtime=nvidia asroy/cuda:10.1-cudnn7-devel-ubuntu18.04-latest /bin/bash