From 041266547d3f8d6a83bd7381fcef0fb8bc559682 Mon Sep 17 00:00:00 2001 From: LJXjean <1372682462@qq.com> Date: Fri, 4 Jul 2025 13:12:49 +0200 Subject: [PATCH 1/6] fix folder name typos --- {doc => docs}/.gitignore | 0 {doc => docs}/README.md | 0 {doc => docs}/benchmark/broadcast.md | 0 {doc => docs}/benchmark/matmul.md | 0 {doc => docs}/benchmark/reduce.md | 0 {doc => docs}/conf.py | 0 {doc => docs}/deepxIR/ir.md | 0 {doc => docs}/deepxIR/readme.md | 0 {doc => docs}/design.md | 0 {doc/excuter => docs/executor}/deepx.op.drawio | 0 {doc/excuter => docs/executor}/deepx.op.drawio.svg | 0 {doc/excuter => docs/executor}/deepx.op.jpg | Bin {doc/excuter => docs/executor}/excuter.md | 0 {doc/excuter => docs/executor}/mix_precision.md | 0 .../executor}/op-mem-cuda/cublas/api.md | 0 .../executor}/op-mem-cuda/cublaslt/api.md | 0 {doc/excuter => docs/executor}/op-mem-cuda/list.md | 0 .../executor}/op-mem-ompsimd/contribute.md | 0 .../executor}/op-mem-ompsimd/list.md | 0 .../executor}/op-mem-ompsimd/range.md | 0 {doc/excuter => docs/executor}/welcome.md | 0 {doc => docs}/front/aboutop.md | 0 {doc => docs}/front/deepx.jpg | Bin {doc => docs}/front/deepx.op.drawio.svg | 0 {doc => docs}/front/deepxpy.drawio.svg | 0 {doc => docs}/front/front.md | 0 {doc => docs}/front/graph.md | 0 {doc => docs}/front/node.md | 0 {doc => docs}/front/op.md | 0 {doc => docs}/front/py/about.md | 0 {doc => docs}/front/py/contribute.md | 0 {doc => docs}/front/py/deepx.rst | 0 {doc => docs}/highway.md | 0 {doc => docs}/index.rst | 0 {doc => docs}/language.md | 0 {doc => docs}/scheduler/scheduler.md | 0 {excuter => executor}/cpp-common/CMakeLists.txt | 0 .../cpp-common/src/client/udpserver.cpp | 0 .../cpp-common/src/client/udpserver.hpp | 0 .../cpp-common/src/client/unixsocketserver.cpp | 0 .../cpp-common/src/client/unixsocketserver.hpp | 0 .../cpp-common/src/client/worker.hpp | 0 .../cpp-common/src/deepx/dtype.hpp | 0 .../cpp-common/src/deepx/mem/mem.hpp | 0 .../cpp-common/src/deepx/shape.cpp | 0 .../cpp-common/src/deepx/shape.hpp | 0 .../cpp-common/src/deepx/shape_changeshape.cpp | 0 .../cpp-common/src/deepx/shape_changeshape.hpp | 0 .../cpp-common/src/deepx/shape_matmul.cpp | 0 .../cpp-common/src/deepx/shape_matmul.hpp | 0 .../cpp-common/src/deepx/shape_range.cpp | 0 .../cpp-common/src/deepx/shape_reduce.cpp | 0 .../cpp-common/src/deepx/shape_reduce.hpp | 0 .../cpp-common/src/deepx/shape_tensorinit.cpp | 0 .../cpp-common/src/deepx/shape_tensorinit.hpp | 0 .../cpp-common/src/deepx/tensor.hpp | 0 .../cpp-common/src/deepx/tensorbase.hpp | 0 .../cpp-common/src/deepx/tensorfunc/authors.hpp | 0 .../cpp-common/src/deepx/tensorfunc/changeshape.hpp | 0 .../cpp-common/src/deepx/tensorfunc/elementwise.hpp | 0 .../cpp-common/src/deepx/tensorfunc/init.hpp | 0 .../cpp-common/src/deepx/tensorfunc/io.hpp | 0 .../cpp-common/src/deepx/tensorfunc/matmul.hpp | 0 .../cpp-common/src/deepx/tensorfunc/reduce.hpp | 0 .../cpp-common/src/deepx/tensorfunc/tensorlife.hpp | 0 .../cpp-common/src/deepx/tf/tf.cpp | 0 .../cpp-common/src/deepx/tf/tf.hpp | 0 .../cpp-common/src/deepx/tf/tffactory.cpp | 0 .../cpp-common/src/deepx/tf/tffactory.hpp | 0 .../cpp-common/src/deepx/vector_combination.cpp | 0 .../cpp-common/src/deepx/vector_combination.hpp | 0 .../cpp-common/src/stdutil/error.hpp | 0 {excuter => executor}/cpp-common/src/stdutil/fs.cpp | 0 {excuter => executor}/cpp-common/src/stdutil/fs.hpp | 0 .../cpp-common/src/stdutil/num.cpp | 0 .../cpp-common/src/stdutil/num.hpp | 0 .../cpp-common/src/stdutil/print.hpp | 0 .../cpp-common/src/stdutil/string.cpp | 0 .../cpp-common/src/stdutil/string.hpp | 0 .../cpp-common/src/stdutil/time.hpp | 0 .../cpp-common/src/stdutil/vector.hpp | 0 {excuter => executor}/cpp-common/test/0_dtypes.cpp | 0 {excuter => executor}/cpp-common/test/1_tf.cpp | 0 {excuter => executor}/cpp-common/test/1_tfcheck.cpp | 0 .../cpp-common/test/2_saveload.cpp | 0 .../cpp-common/test/CMakeLists.txt | 0 {excuter => executor}/op-mem-cuda/.gitignore | 0 {excuter => executor}/op-mem-cuda/CMakeLists.txt | 0 {excuter => executor}/op-mem-cuda/Dockerfile | 0 {excuter => executor}/op-mem-cuda/README.md | 0 {excuter => executor}/op-mem-cuda/build.sh | 0 .../op-mem-cuda/doc/00_quickstart.md | 0 {excuter => executor}/op-mem-cuda/doc/01_layout.md | 0 .../op-mem-cuda/doc/02_layout_algebra.md | 0 {excuter => executor}/op-mem-cuda/doc/03_tensor.md | 0 .../op-mem-cuda/doc/04_algorithms.md | 0 .../op-mem-cuda/doc/0t_mma_atom.md | 0 .../op-mem-cuda/doc/0x_gemm_tutorial.md | 0 .../op-mem-cuda/doc/0y_predication.md | 0 .../op-mem-cuda/doc/0z_tma_tensors.md | 0 {excuter => executor}/op-mem-cuda/dockerbuild.sh | 0 {excuter => executor}/op-mem-cuda/log.md | 0 .../op-mem-cuda/src/client/main.cpp | 0 .../op-mem-cuda/src/client/tfs.cpp | 0 .../op-mem-cuda/src/client/tfs.hpp | 0 .../op-mem-cuda/src/deepx/dtype_cuda.hpp | 0 .../op-mem-cuda/src/deepx/mem/mem_cuda.hpp | 0 .../src/deepx/tensorfunc/changeshape_miaobyte.cu | 0 .../src/deepx/tensorfunc/changeshape_miaobyte.cuh | 0 .../src/deepx/tensorfunc/changeshape_miaobyte.hpp | 0 .../op-mem-cuda/src/deepx/tensorfunc/cuda.hpp | 0 .../src/deepx/tensorfunc/cuda_atomic.cuh | 0 .../op-mem-cuda/src/deepx/tensorfunc/cuda_math.cuh | 0 .../deepx/tensorfunc/elementwise_cublas_basic.hpp | 0 .../deepx/tensorfunc/elementwise_miaobyte_basic.cu | 0 .../deepx/tensorfunc/elementwise_miaobyte_basic.cuh | 0 .../deepx/tensorfunc/elementwise_miaobyte_basic.hpp | 0 .../tensorfunc/elementwise_miaobyte_compare.cu | 0 .../tensorfunc/elementwise_miaobyte_compare.cuh | 0 .../tensorfunc/elementwise_miaobyte_compare.hpp | 0 .../deepx/tensorfunc/elementwise_miaobyte_sin.cu | 0 .../deepx/tensorfunc/elementwise_miaobyte_sin.cuh | 0 .../deepx/tensorfunc/elementwise_miaobyte_sin.hpp | 0 .../deepx/tensorfunc/elementwise_miaobyte_sqrt.cu | 0 .../deepx/tensorfunc/elementwise_miaobyte_sqrt.cuh | 0 .../deepx/tensorfunc/elementwise_miaobyte_sqrt.hpp | 0 .../src/deepx/tensorfunc/init_miaobyte.cu | 0 .../src/deepx/tensorfunc/init_miaobyte.cuh | 0 .../src/deepx/tensorfunc/init_miaobyte.hpp | 0 .../src/deepx/tensorfunc/io_miaobyte.hpp | 0 .../src/deepx/tensorfunc/matmul_cublas.hpp | 0 .../src/deepx/tensorfunc/new_mempool.hpp | 0 .../src/deepx/tensorfunc/reduce_miaobyte.cu | 0 .../src/deepx/tensorfunc/reduce_miaobyte.cuh | 0 .../src/deepx/tensorfunc/reduce_miaobyte.hpp | 0 .../src/deepx/tensorfunc/tensor_cuda.cuh | 0 .../src/deepx/tensorfunc/tensorlife_miaobyte.hpp | 0 .../src/deepx/tensorfunc/vector_cuda.cuh | 0 .../op-mem-cuda/src/deepx/tf/arg.hpp | 0 .../op-mem-cuda/src/deepx/tf/changeshape.hpp | 0 .../op-mem-cuda/src/deepx/tf/elementwise_basic.hpp | 0 .../src/deepx/tf/elementwise_compare.hpp | 0 .../op-mem-cuda/src/deepx/tf/elementwise_sin.hpp | 0 .../op-mem-cuda/src/deepx/tf/elementwise_sqrt.hpp | 0 .../op-mem-cuda/src/deepx/tf/init.hpp | 0 .../op-mem-cuda/src/deepx/tf/io.hpp | 0 .../op-mem-cuda/src/deepx/tf/matmul.hpp | 0 .../op-mem-cuda/src/deepx/tf/reduce.hpp | 0 .../op-mem-cuda/src/deepx/tf/tensorlife.hpp | 0 .../op-mem-cuda/test/op/CMakeLists.txt | 0 .../op-mem-cuda/test/tensorfunc/0_new.cpp | 0 .../op-mem-cuda/test/tensorfunc/1_cublas_add.cpp | 0 .../op-mem-cuda/test/tensorfunc/1_cublas_matmul.cpp | 0 .../op-mem-cuda/test/tensorfunc/2_changeshape.cpp | 0 .../op-mem-cuda/test/tensorfunc/CMakeLists.txt | 0 {excuter => executor}/op-mem-mps/swift.md | 0 {excuter => executor}/op-mem-ompsimd/.cursorignore | 0 {excuter => executor}/op-mem-ompsimd/.cursorrules | 0 {excuter => executor}/op-mem-ompsimd/.gitignore | 0 {excuter => executor}/op-mem-ompsimd/CMakeLists.txt | 0 {excuter => executor}/op-mem-ompsimd/Dockerfile | 0 {excuter => executor}/op-mem-ompsimd/dockerbuild.sh | 0 {excuter => executor}/op-mem-ompsimd/log.md | 0 .../op-mem-ompsimd/src/client/main.cpp | 0 .../op-mem-ompsimd/src/client/tfs.cpp | 0 .../op-mem-ompsimd/src/client/tfs.hpp | 0 .../op-mem-ompsimd/src/deepx/dtype_ompsimd.hpp | 0 .../op-mem-ompsimd/src/deepx/mem/mem_ompsimd.hpp | 0 .../src/deepx/tensorfunc/changeshape_miaobyte.hpp | 0 .../src/deepx/tensorfunc/elementwise_cblas.hpp | 0 .../src/deepx/tensorfunc/elementwise_miaobyte.hpp | 0 .../op-mem-ompsimd/src/deepx/tensorfunc/equal.hpp | 0 .../op-mem-ompsimd/src/deepx/tensorfunc/highway.hpp | 0 .../src/deepx/tensorfunc/init_miaobyte.hpp | 0 .../src/deepx/tensorfunc/io_miaobyte.hpp | 0 .../src/deepx/tensorfunc/matmul_cblas.hpp | 0 .../src/deepx/tensorfunc/matmul_miaobyte.hpp | 0 .../src/deepx/tensorfunc/new_mempool.hpp | 0 .../src/deepx/tensorfunc/reduce_miaobyte.hpp | 0 .../src/deepx/tensorfunc/tensorlife_miaobyte.hpp | 0 .../op-mem-ompsimd/src/deepx/tf/arg.hpp | 0 .../op-mem-ompsimd/src/deepx/tf/changeshape.hpp | 0 .../op-mem-ompsimd/src/deepx/tf/elementwise.hpp | 0 .../op-mem-ompsimd/src/deepx/tf/init.hpp | 0 .../op-mem-ompsimd/src/deepx/tf/io.hpp | 0 .../op-mem-ompsimd/src/deepx/tf/matmul.hpp | 0 .../op-mem-ompsimd/src/deepx/tf/reduce.hpp | 0 .../op-mem-ompsimd/src/deepx/tf/tensorlife.hpp | 0 .../op-mem-ompsimd/test/op/1_mem.cpp | 0 .../op-mem-ompsimd/test/op/CMakeLists.txt | 0 .../op-mem-ompsimd/test/tensorfunc/1_shape.cpp | 0 .../test/tensorfunc/2_shape_combintion.cpp | 0 .../test/tensorfunc/2_tensor_equal.cpp | 0 .../op-mem-ompsimd/test/tensorfunc/2_tensor_new.cpp | 0 .../test/tensorfunc/2_tensor_range.cpp | 0 .../test/tensorfunc/2_tensor_range.py | 0 .../test/tensorfunc/3_tensor_print.cpp | 0 .../op-mem-ompsimd/test/tensorfunc/4_tensor_add.cpp | 0 .../test/tensorfunc/4_tensor_matmul.cpp | 0 .../op-mem-ompsimd/test/tensorfunc/4_tensor_max.cpp | 0 .../op-mem-ompsimd/test/tensorfunc/4_tensor_mul.cpp | 0 .../op-mem-ompsimd/test/tensorfunc/4_tensor_sub.cpp | 0 .../op-mem-ompsimd/test/tensorfunc/5_tensor_sum.cpp | 0 .../test/tensorfunc/6_tensor_broadcast.cpp | 0 .../test/tensorfunc/7_tensor_transpose.cpp | 0 .../test/tensorfunc/8_tensor_concat.cpp | 0 .../op-mem-ompsimd/test/tensorfunc/CMakeLists.txt | 0 .../op-mem-ompsimd/test/tensorfunc/tensorutil.hpp | 0 208 files changed, 0 insertions(+), 0 deletions(-) rename {doc => docs}/.gitignore (100%) rename {doc => docs}/README.md (100%) rename {doc => docs}/benchmark/broadcast.md (100%) rename {doc => docs}/benchmark/matmul.md (100%) rename {doc => docs}/benchmark/reduce.md (100%) rename {doc => docs}/conf.py (100%) rename {doc => docs}/deepxIR/ir.md (100%) rename {doc => docs}/deepxIR/readme.md (100%) rename {doc => docs}/design.md (100%) rename {doc/excuter => docs/executor}/deepx.op.drawio (100%) rename {doc/excuter => docs/executor}/deepx.op.drawio.svg (100%) rename {doc/excuter => docs/executor}/deepx.op.jpg (100%) rename {doc/excuter => docs/executor}/excuter.md (100%) rename {doc/excuter => docs/executor}/mix_precision.md (100%) rename {doc/excuter => docs/executor}/op-mem-cuda/cublas/api.md (100%) rename {doc/excuter => docs/executor}/op-mem-cuda/cublaslt/api.md (100%) rename {doc/excuter => docs/executor}/op-mem-cuda/list.md (100%) rename {doc/excuter => docs/executor}/op-mem-ompsimd/contribute.md (100%) rename {doc/excuter => docs/executor}/op-mem-ompsimd/list.md (100%) rename {doc/excuter => docs/executor}/op-mem-ompsimd/range.md (100%) rename {doc/excuter => docs/executor}/welcome.md (100%) rename {doc => docs}/front/aboutop.md (100%) rename {doc => docs}/front/deepx.jpg (100%) rename {doc => docs}/front/deepx.op.drawio.svg (100%) rename {doc => docs}/front/deepxpy.drawio.svg (100%) rename {doc => docs}/front/front.md (100%) rename {doc => docs}/front/graph.md (100%) rename {doc => docs}/front/node.md (100%) rename {doc => docs}/front/op.md (100%) rename {doc => docs}/front/py/about.md (100%) rename {doc => docs}/front/py/contribute.md (100%) rename {doc => docs}/front/py/deepx.rst (100%) rename {doc => docs}/highway.md (100%) rename {doc => docs}/index.rst (100%) rename {doc => docs}/language.md (100%) rename {doc => docs}/scheduler/scheduler.md (100%) rename {excuter => executor}/cpp-common/CMakeLists.txt (100%) rename {excuter => executor}/cpp-common/src/client/udpserver.cpp (100%) rename {excuter => executor}/cpp-common/src/client/udpserver.hpp (100%) rename {excuter => executor}/cpp-common/src/client/unixsocketserver.cpp (100%) rename {excuter => executor}/cpp-common/src/client/unixsocketserver.hpp (100%) rename {excuter => executor}/cpp-common/src/client/worker.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/dtype.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/mem/mem.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/shape.cpp (100%) rename {excuter => executor}/cpp-common/src/deepx/shape.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/shape_changeshape.cpp (100%) rename {excuter => executor}/cpp-common/src/deepx/shape_changeshape.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/shape_matmul.cpp (100%) rename {excuter => executor}/cpp-common/src/deepx/shape_matmul.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/shape_range.cpp (100%) rename {excuter => executor}/cpp-common/src/deepx/shape_reduce.cpp (100%) rename {excuter => executor}/cpp-common/src/deepx/shape_reduce.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/shape_tensorinit.cpp (100%) rename {excuter => executor}/cpp-common/src/deepx/shape_tensorinit.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/tensor.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/tensorbase.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/tensorfunc/authors.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/tensorfunc/changeshape.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/tensorfunc/elementwise.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/tensorfunc/init.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/tensorfunc/io.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/tensorfunc/matmul.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/tensorfunc/reduce.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/tensorfunc/tensorlife.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/tf/tf.cpp (100%) rename {excuter => executor}/cpp-common/src/deepx/tf/tf.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/tf/tffactory.cpp (100%) rename {excuter => executor}/cpp-common/src/deepx/tf/tffactory.hpp (100%) rename {excuter => executor}/cpp-common/src/deepx/vector_combination.cpp (100%) rename {excuter => executor}/cpp-common/src/deepx/vector_combination.hpp (100%) rename {excuter => executor}/cpp-common/src/stdutil/error.hpp (100%) rename {excuter => executor}/cpp-common/src/stdutil/fs.cpp (100%) rename {excuter => executor}/cpp-common/src/stdutil/fs.hpp (100%) rename {excuter => executor}/cpp-common/src/stdutil/num.cpp (100%) rename {excuter => executor}/cpp-common/src/stdutil/num.hpp (100%) rename {excuter => executor}/cpp-common/src/stdutil/print.hpp (100%) rename {excuter => executor}/cpp-common/src/stdutil/string.cpp (100%) rename {excuter => executor}/cpp-common/src/stdutil/string.hpp (100%) rename {excuter => executor}/cpp-common/src/stdutil/time.hpp (100%) rename {excuter => executor}/cpp-common/src/stdutil/vector.hpp (100%) rename {excuter => executor}/cpp-common/test/0_dtypes.cpp (100%) rename {excuter => executor}/cpp-common/test/1_tf.cpp (100%) rename {excuter => executor}/cpp-common/test/1_tfcheck.cpp (100%) rename {excuter => executor}/cpp-common/test/2_saveload.cpp (100%) rename {excuter => executor}/cpp-common/test/CMakeLists.txt (100%) rename {excuter => executor}/op-mem-cuda/.gitignore (100%) rename {excuter => executor}/op-mem-cuda/CMakeLists.txt (100%) rename {excuter => executor}/op-mem-cuda/Dockerfile (100%) rename {excuter => executor}/op-mem-cuda/README.md (100%) rename {excuter => executor}/op-mem-cuda/build.sh (100%) rename {excuter => executor}/op-mem-cuda/doc/00_quickstart.md (100%) rename {excuter => executor}/op-mem-cuda/doc/01_layout.md (100%) rename {excuter => executor}/op-mem-cuda/doc/02_layout_algebra.md (100%) rename {excuter => executor}/op-mem-cuda/doc/03_tensor.md (100%) rename {excuter => executor}/op-mem-cuda/doc/04_algorithms.md (100%) rename {excuter => executor}/op-mem-cuda/doc/0t_mma_atom.md (100%) rename {excuter => executor}/op-mem-cuda/doc/0x_gemm_tutorial.md (100%) rename {excuter => executor}/op-mem-cuda/doc/0y_predication.md (100%) rename {excuter => executor}/op-mem-cuda/doc/0z_tma_tensors.md (100%) rename {excuter => executor}/op-mem-cuda/dockerbuild.sh (100%) rename {excuter => executor}/op-mem-cuda/log.md (100%) rename {excuter => executor}/op-mem-cuda/src/client/main.cpp (100%) rename {excuter => executor}/op-mem-cuda/src/client/tfs.cpp (100%) rename {excuter => executor}/op-mem-cuda/src/client/tfs.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/dtype_cuda.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/mem/mem_cuda.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/changeshape_miaobyte.cu (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/changeshape_miaobyte.cuh (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/changeshape_miaobyte.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/cuda.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/cuda_atomic.cuh (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/cuda_math.cuh (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/elementwise_cublas_basic.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_basic.cu (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_basic.cuh (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_basic.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_compare.cu (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_compare.cuh (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_compare.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sin.cu (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sin.cuh (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sin.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sqrt.cu (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sqrt.cuh (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sqrt.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.cu (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.cuh (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/io_miaobyte.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/matmul_cublas.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/new_mempool.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/reduce_miaobyte.cu (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/reduce_miaobyte.cuh (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/reduce_miaobyte.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/tensor_cuda.cuh (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/tensorlife_miaobyte.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tensorfunc/vector_cuda.cuh (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tf/arg.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tf/changeshape.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tf/elementwise_basic.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tf/elementwise_compare.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tf/elementwise_sin.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tf/elementwise_sqrt.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tf/init.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tf/io.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tf/matmul.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tf/reduce.hpp (100%) rename {excuter => executor}/op-mem-cuda/src/deepx/tf/tensorlife.hpp (100%) rename {excuter => executor}/op-mem-cuda/test/op/CMakeLists.txt (100%) rename {excuter => executor}/op-mem-cuda/test/tensorfunc/0_new.cpp (100%) rename {excuter => executor}/op-mem-cuda/test/tensorfunc/1_cublas_add.cpp (100%) rename {excuter => executor}/op-mem-cuda/test/tensorfunc/1_cublas_matmul.cpp (100%) rename {excuter => executor}/op-mem-cuda/test/tensorfunc/2_changeshape.cpp (100%) rename {excuter => executor}/op-mem-cuda/test/tensorfunc/CMakeLists.txt (100%) rename {excuter => executor}/op-mem-mps/swift.md (100%) rename {excuter => executor}/op-mem-ompsimd/.cursorignore (100%) rename {excuter => executor}/op-mem-ompsimd/.cursorrules (100%) rename {excuter => executor}/op-mem-ompsimd/.gitignore (100%) rename {excuter => executor}/op-mem-ompsimd/CMakeLists.txt (100%) rename {excuter => executor}/op-mem-ompsimd/Dockerfile (100%) rename {excuter => executor}/op-mem-ompsimd/dockerbuild.sh (100%) rename {excuter => executor}/op-mem-ompsimd/log.md (100%) rename {excuter => executor}/op-mem-ompsimd/src/client/main.cpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/client/tfs.cpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/client/tfs.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/dtype_ompsimd.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/mem/mem_ompsimd.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tensorfunc/changeshape_miaobyte.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tensorfunc/elementwise_cblas.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tensorfunc/elementwise_miaobyte.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tensorfunc/equal.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tensorfunc/highway.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tensorfunc/init_miaobyte.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tensorfunc/io_miaobyte.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tensorfunc/matmul_cblas.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tensorfunc/matmul_miaobyte.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tensorfunc/new_mempool.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tensorfunc/reduce_miaobyte.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tensorfunc/tensorlife_miaobyte.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tf/arg.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tf/changeshape.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tf/elementwise.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tf/init.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tf/io.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tf/matmul.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tf/reduce.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/src/deepx/tf/tensorlife.hpp (100%) rename {excuter => executor}/op-mem-ompsimd/test/op/1_mem.cpp (100%) rename {excuter => executor}/op-mem-ompsimd/test/op/CMakeLists.txt (100%) rename {excuter => executor}/op-mem-ompsimd/test/tensorfunc/1_shape.cpp (100%) rename {excuter => executor}/op-mem-ompsimd/test/tensorfunc/2_shape_combintion.cpp (100%) rename {excuter => executor}/op-mem-ompsimd/test/tensorfunc/2_tensor_equal.cpp (100%) rename {excuter => executor}/op-mem-ompsimd/test/tensorfunc/2_tensor_new.cpp (100%) rename {excuter => executor}/op-mem-ompsimd/test/tensorfunc/2_tensor_range.cpp (100%) rename {excuter => executor}/op-mem-ompsimd/test/tensorfunc/2_tensor_range.py (100%) rename {excuter => executor}/op-mem-ompsimd/test/tensorfunc/3_tensor_print.cpp (100%) rename {excuter => executor}/op-mem-ompsimd/test/tensorfunc/4_tensor_add.cpp (100%) rename {excuter => executor}/op-mem-ompsimd/test/tensorfunc/4_tensor_matmul.cpp (100%) rename {excuter => executor}/op-mem-ompsimd/test/tensorfunc/4_tensor_max.cpp (100%) rename {excuter => executor}/op-mem-ompsimd/test/tensorfunc/4_tensor_mul.cpp (100%) rename {excuter => executor}/op-mem-ompsimd/test/tensorfunc/4_tensor_sub.cpp (100%) rename {excuter => executor}/op-mem-ompsimd/test/tensorfunc/5_tensor_sum.cpp (100%) rename {excuter => executor}/op-mem-ompsimd/test/tensorfunc/6_tensor_broadcast.cpp (100%) rename {excuter => executor}/op-mem-ompsimd/test/tensorfunc/7_tensor_transpose.cpp (100%) rename {excuter => executor}/op-mem-ompsimd/test/tensorfunc/8_tensor_concat.cpp (100%) rename {excuter => executor}/op-mem-ompsimd/test/tensorfunc/CMakeLists.txt (100%) rename {excuter => executor}/op-mem-ompsimd/test/tensorfunc/tensorutil.hpp (100%) diff --git a/doc/.gitignore b/docs/.gitignore similarity index 100% rename from doc/.gitignore rename to docs/.gitignore diff --git a/doc/README.md b/docs/README.md similarity index 100% rename from doc/README.md rename to docs/README.md diff --git a/doc/benchmark/broadcast.md b/docs/benchmark/broadcast.md similarity index 100% rename from doc/benchmark/broadcast.md rename to docs/benchmark/broadcast.md diff --git a/doc/benchmark/matmul.md b/docs/benchmark/matmul.md similarity index 100% rename from doc/benchmark/matmul.md rename to docs/benchmark/matmul.md diff --git a/doc/benchmark/reduce.md b/docs/benchmark/reduce.md similarity index 100% rename from doc/benchmark/reduce.md rename to docs/benchmark/reduce.md diff --git a/doc/conf.py b/docs/conf.py similarity index 100% rename from doc/conf.py rename to docs/conf.py diff --git a/doc/deepxIR/ir.md b/docs/deepxIR/ir.md similarity index 100% rename from doc/deepxIR/ir.md rename to docs/deepxIR/ir.md diff --git a/doc/deepxIR/readme.md b/docs/deepxIR/readme.md similarity index 100% rename from doc/deepxIR/readme.md rename to docs/deepxIR/readme.md diff --git a/doc/design.md b/docs/design.md similarity index 100% rename from doc/design.md rename to docs/design.md diff --git a/doc/excuter/deepx.op.drawio b/docs/executor/deepx.op.drawio similarity index 100% rename from doc/excuter/deepx.op.drawio rename to docs/executor/deepx.op.drawio diff --git a/doc/excuter/deepx.op.drawio.svg b/docs/executor/deepx.op.drawio.svg similarity index 100% rename from doc/excuter/deepx.op.drawio.svg rename to docs/executor/deepx.op.drawio.svg diff --git a/doc/excuter/deepx.op.jpg b/docs/executor/deepx.op.jpg similarity index 100% rename from doc/excuter/deepx.op.jpg rename to docs/executor/deepx.op.jpg diff --git a/doc/excuter/excuter.md b/docs/executor/excuter.md similarity index 100% rename from doc/excuter/excuter.md rename to docs/executor/excuter.md diff --git a/doc/excuter/mix_precision.md b/docs/executor/mix_precision.md similarity index 100% rename from doc/excuter/mix_precision.md rename to docs/executor/mix_precision.md diff --git a/doc/excuter/op-mem-cuda/cublas/api.md b/docs/executor/op-mem-cuda/cublas/api.md similarity index 100% rename from doc/excuter/op-mem-cuda/cublas/api.md rename to docs/executor/op-mem-cuda/cublas/api.md diff --git a/doc/excuter/op-mem-cuda/cublaslt/api.md b/docs/executor/op-mem-cuda/cublaslt/api.md similarity index 100% rename from doc/excuter/op-mem-cuda/cublaslt/api.md rename to docs/executor/op-mem-cuda/cublaslt/api.md diff --git a/doc/excuter/op-mem-cuda/list.md b/docs/executor/op-mem-cuda/list.md similarity index 100% rename from doc/excuter/op-mem-cuda/list.md rename to docs/executor/op-mem-cuda/list.md diff --git a/doc/excuter/op-mem-ompsimd/contribute.md b/docs/executor/op-mem-ompsimd/contribute.md similarity index 100% rename from doc/excuter/op-mem-ompsimd/contribute.md rename to docs/executor/op-mem-ompsimd/contribute.md diff --git a/doc/excuter/op-mem-ompsimd/list.md b/docs/executor/op-mem-ompsimd/list.md similarity index 100% rename from doc/excuter/op-mem-ompsimd/list.md rename to docs/executor/op-mem-ompsimd/list.md diff --git a/doc/excuter/op-mem-ompsimd/range.md b/docs/executor/op-mem-ompsimd/range.md similarity index 100% rename from doc/excuter/op-mem-ompsimd/range.md rename to docs/executor/op-mem-ompsimd/range.md diff --git a/doc/excuter/welcome.md b/docs/executor/welcome.md similarity index 100% rename from doc/excuter/welcome.md rename to docs/executor/welcome.md diff --git a/doc/front/aboutop.md b/docs/front/aboutop.md similarity index 100% rename from doc/front/aboutop.md rename to docs/front/aboutop.md diff --git a/doc/front/deepx.jpg b/docs/front/deepx.jpg similarity index 100% rename from doc/front/deepx.jpg rename to docs/front/deepx.jpg diff --git a/doc/front/deepx.op.drawio.svg b/docs/front/deepx.op.drawio.svg similarity index 100% rename from doc/front/deepx.op.drawio.svg rename to docs/front/deepx.op.drawio.svg diff --git a/doc/front/deepxpy.drawio.svg b/docs/front/deepxpy.drawio.svg similarity index 100% rename from doc/front/deepxpy.drawio.svg rename to docs/front/deepxpy.drawio.svg diff --git a/doc/front/front.md b/docs/front/front.md similarity index 100% rename from doc/front/front.md rename to docs/front/front.md diff --git a/doc/front/graph.md b/docs/front/graph.md similarity index 100% rename from doc/front/graph.md rename to docs/front/graph.md diff --git a/doc/front/node.md b/docs/front/node.md similarity index 100% rename from doc/front/node.md rename to docs/front/node.md diff --git a/doc/front/op.md b/docs/front/op.md similarity index 100% rename from doc/front/op.md rename to docs/front/op.md diff --git a/doc/front/py/about.md b/docs/front/py/about.md similarity index 100% rename from doc/front/py/about.md rename to docs/front/py/about.md diff --git a/doc/front/py/contribute.md b/docs/front/py/contribute.md similarity index 100% rename from doc/front/py/contribute.md rename to docs/front/py/contribute.md diff --git a/doc/front/py/deepx.rst b/docs/front/py/deepx.rst similarity index 100% rename from doc/front/py/deepx.rst rename to docs/front/py/deepx.rst diff --git a/doc/highway.md b/docs/highway.md similarity index 100% rename from doc/highway.md rename to docs/highway.md diff --git a/doc/index.rst b/docs/index.rst similarity index 100% rename from doc/index.rst rename to docs/index.rst diff --git a/doc/language.md b/docs/language.md similarity index 100% rename from doc/language.md rename to docs/language.md diff --git a/doc/scheduler/scheduler.md b/docs/scheduler/scheduler.md similarity index 100% rename from doc/scheduler/scheduler.md rename to docs/scheduler/scheduler.md diff --git a/excuter/cpp-common/CMakeLists.txt b/executor/cpp-common/CMakeLists.txt similarity index 100% rename from excuter/cpp-common/CMakeLists.txt rename to executor/cpp-common/CMakeLists.txt diff --git a/excuter/cpp-common/src/client/udpserver.cpp b/executor/cpp-common/src/client/udpserver.cpp similarity index 100% rename from excuter/cpp-common/src/client/udpserver.cpp rename to executor/cpp-common/src/client/udpserver.cpp diff --git a/excuter/cpp-common/src/client/udpserver.hpp b/executor/cpp-common/src/client/udpserver.hpp similarity index 100% rename from excuter/cpp-common/src/client/udpserver.hpp rename to executor/cpp-common/src/client/udpserver.hpp diff --git a/excuter/cpp-common/src/client/unixsocketserver.cpp b/executor/cpp-common/src/client/unixsocketserver.cpp similarity index 100% rename from excuter/cpp-common/src/client/unixsocketserver.cpp rename to executor/cpp-common/src/client/unixsocketserver.cpp diff --git a/excuter/cpp-common/src/client/unixsocketserver.hpp b/executor/cpp-common/src/client/unixsocketserver.hpp similarity index 100% rename from excuter/cpp-common/src/client/unixsocketserver.hpp rename to executor/cpp-common/src/client/unixsocketserver.hpp diff --git a/excuter/cpp-common/src/client/worker.hpp b/executor/cpp-common/src/client/worker.hpp similarity index 100% rename from excuter/cpp-common/src/client/worker.hpp rename to executor/cpp-common/src/client/worker.hpp diff --git a/excuter/cpp-common/src/deepx/dtype.hpp b/executor/cpp-common/src/deepx/dtype.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/dtype.hpp rename to executor/cpp-common/src/deepx/dtype.hpp diff --git a/excuter/cpp-common/src/deepx/mem/mem.hpp b/executor/cpp-common/src/deepx/mem/mem.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/mem/mem.hpp rename to executor/cpp-common/src/deepx/mem/mem.hpp diff --git a/excuter/cpp-common/src/deepx/shape.cpp b/executor/cpp-common/src/deepx/shape.cpp similarity index 100% rename from excuter/cpp-common/src/deepx/shape.cpp rename to executor/cpp-common/src/deepx/shape.cpp diff --git a/excuter/cpp-common/src/deepx/shape.hpp b/executor/cpp-common/src/deepx/shape.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/shape.hpp rename to executor/cpp-common/src/deepx/shape.hpp diff --git a/excuter/cpp-common/src/deepx/shape_changeshape.cpp b/executor/cpp-common/src/deepx/shape_changeshape.cpp similarity index 100% rename from excuter/cpp-common/src/deepx/shape_changeshape.cpp rename to executor/cpp-common/src/deepx/shape_changeshape.cpp diff --git a/excuter/cpp-common/src/deepx/shape_changeshape.hpp b/executor/cpp-common/src/deepx/shape_changeshape.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/shape_changeshape.hpp rename to executor/cpp-common/src/deepx/shape_changeshape.hpp diff --git a/excuter/cpp-common/src/deepx/shape_matmul.cpp b/executor/cpp-common/src/deepx/shape_matmul.cpp similarity index 100% rename from excuter/cpp-common/src/deepx/shape_matmul.cpp rename to executor/cpp-common/src/deepx/shape_matmul.cpp diff --git a/excuter/cpp-common/src/deepx/shape_matmul.hpp b/executor/cpp-common/src/deepx/shape_matmul.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/shape_matmul.hpp rename to executor/cpp-common/src/deepx/shape_matmul.hpp diff --git a/excuter/cpp-common/src/deepx/shape_range.cpp b/executor/cpp-common/src/deepx/shape_range.cpp similarity index 100% rename from excuter/cpp-common/src/deepx/shape_range.cpp rename to executor/cpp-common/src/deepx/shape_range.cpp diff --git a/excuter/cpp-common/src/deepx/shape_reduce.cpp b/executor/cpp-common/src/deepx/shape_reduce.cpp similarity index 100% rename from excuter/cpp-common/src/deepx/shape_reduce.cpp rename to executor/cpp-common/src/deepx/shape_reduce.cpp diff --git a/excuter/cpp-common/src/deepx/shape_reduce.hpp b/executor/cpp-common/src/deepx/shape_reduce.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/shape_reduce.hpp rename to executor/cpp-common/src/deepx/shape_reduce.hpp diff --git a/excuter/cpp-common/src/deepx/shape_tensorinit.cpp b/executor/cpp-common/src/deepx/shape_tensorinit.cpp similarity index 100% rename from excuter/cpp-common/src/deepx/shape_tensorinit.cpp rename to executor/cpp-common/src/deepx/shape_tensorinit.cpp diff --git a/excuter/cpp-common/src/deepx/shape_tensorinit.hpp b/executor/cpp-common/src/deepx/shape_tensorinit.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/shape_tensorinit.hpp rename to executor/cpp-common/src/deepx/shape_tensorinit.hpp diff --git a/excuter/cpp-common/src/deepx/tensor.hpp b/executor/cpp-common/src/deepx/tensor.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/tensor.hpp rename to executor/cpp-common/src/deepx/tensor.hpp diff --git a/excuter/cpp-common/src/deepx/tensorbase.hpp b/executor/cpp-common/src/deepx/tensorbase.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/tensorbase.hpp rename to executor/cpp-common/src/deepx/tensorbase.hpp diff --git a/excuter/cpp-common/src/deepx/tensorfunc/authors.hpp b/executor/cpp-common/src/deepx/tensorfunc/authors.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/tensorfunc/authors.hpp rename to executor/cpp-common/src/deepx/tensorfunc/authors.hpp diff --git a/excuter/cpp-common/src/deepx/tensorfunc/changeshape.hpp b/executor/cpp-common/src/deepx/tensorfunc/changeshape.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/tensorfunc/changeshape.hpp rename to executor/cpp-common/src/deepx/tensorfunc/changeshape.hpp diff --git a/excuter/cpp-common/src/deepx/tensorfunc/elementwise.hpp b/executor/cpp-common/src/deepx/tensorfunc/elementwise.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/tensorfunc/elementwise.hpp rename to executor/cpp-common/src/deepx/tensorfunc/elementwise.hpp diff --git a/excuter/cpp-common/src/deepx/tensorfunc/init.hpp b/executor/cpp-common/src/deepx/tensorfunc/init.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/tensorfunc/init.hpp rename to executor/cpp-common/src/deepx/tensorfunc/init.hpp diff --git a/excuter/cpp-common/src/deepx/tensorfunc/io.hpp b/executor/cpp-common/src/deepx/tensorfunc/io.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/tensorfunc/io.hpp rename to executor/cpp-common/src/deepx/tensorfunc/io.hpp diff --git a/excuter/cpp-common/src/deepx/tensorfunc/matmul.hpp b/executor/cpp-common/src/deepx/tensorfunc/matmul.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/tensorfunc/matmul.hpp rename to executor/cpp-common/src/deepx/tensorfunc/matmul.hpp diff --git a/excuter/cpp-common/src/deepx/tensorfunc/reduce.hpp b/executor/cpp-common/src/deepx/tensorfunc/reduce.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/tensorfunc/reduce.hpp rename to executor/cpp-common/src/deepx/tensorfunc/reduce.hpp diff --git a/excuter/cpp-common/src/deepx/tensorfunc/tensorlife.hpp b/executor/cpp-common/src/deepx/tensorfunc/tensorlife.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/tensorfunc/tensorlife.hpp rename to executor/cpp-common/src/deepx/tensorfunc/tensorlife.hpp diff --git a/excuter/cpp-common/src/deepx/tf/tf.cpp b/executor/cpp-common/src/deepx/tf/tf.cpp similarity index 100% rename from excuter/cpp-common/src/deepx/tf/tf.cpp rename to executor/cpp-common/src/deepx/tf/tf.cpp diff --git a/excuter/cpp-common/src/deepx/tf/tf.hpp b/executor/cpp-common/src/deepx/tf/tf.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/tf/tf.hpp rename to executor/cpp-common/src/deepx/tf/tf.hpp diff --git a/excuter/cpp-common/src/deepx/tf/tffactory.cpp b/executor/cpp-common/src/deepx/tf/tffactory.cpp similarity index 100% rename from excuter/cpp-common/src/deepx/tf/tffactory.cpp rename to executor/cpp-common/src/deepx/tf/tffactory.cpp diff --git a/excuter/cpp-common/src/deepx/tf/tffactory.hpp b/executor/cpp-common/src/deepx/tf/tffactory.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/tf/tffactory.hpp rename to executor/cpp-common/src/deepx/tf/tffactory.hpp diff --git a/excuter/cpp-common/src/deepx/vector_combination.cpp b/executor/cpp-common/src/deepx/vector_combination.cpp similarity index 100% rename from excuter/cpp-common/src/deepx/vector_combination.cpp rename to executor/cpp-common/src/deepx/vector_combination.cpp diff --git a/excuter/cpp-common/src/deepx/vector_combination.hpp b/executor/cpp-common/src/deepx/vector_combination.hpp similarity index 100% rename from excuter/cpp-common/src/deepx/vector_combination.hpp rename to executor/cpp-common/src/deepx/vector_combination.hpp diff --git a/excuter/cpp-common/src/stdutil/error.hpp b/executor/cpp-common/src/stdutil/error.hpp similarity index 100% rename from excuter/cpp-common/src/stdutil/error.hpp rename to executor/cpp-common/src/stdutil/error.hpp diff --git a/excuter/cpp-common/src/stdutil/fs.cpp b/executor/cpp-common/src/stdutil/fs.cpp similarity index 100% rename from excuter/cpp-common/src/stdutil/fs.cpp rename to executor/cpp-common/src/stdutil/fs.cpp diff --git a/excuter/cpp-common/src/stdutil/fs.hpp b/executor/cpp-common/src/stdutil/fs.hpp similarity index 100% rename from excuter/cpp-common/src/stdutil/fs.hpp rename to executor/cpp-common/src/stdutil/fs.hpp diff --git a/excuter/cpp-common/src/stdutil/num.cpp b/executor/cpp-common/src/stdutil/num.cpp similarity index 100% rename from excuter/cpp-common/src/stdutil/num.cpp rename to executor/cpp-common/src/stdutil/num.cpp diff --git a/excuter/cpp-common/src/stdutil/num.hpp b/executor/cpp-common/src/stdutil/num.hpp similarity index 100% rename from excuter/cpp-common/src/stdutil/num.hpp rename to executor/cpp-common/src/stdutil/num.hpp diff --git a/excuter/cpp-common/src/stdutil/print.hpp b/executor/cpp-common/src/stdutil/print.hpp similarity index 100% rename from excuter/cpp-common/src/stdutil/print.hpp rename to executor/cpp-common/src/stdutil/print.hpp diff --git a/excuter/cpp-common/src/stdutil/string.cpp b/executor/cpp-common/src/stdutil/string.cpp similarity index 100% rename from excuter/cpp-common/src/stdutil/string.cpp rename to executor/cpp-common/src/stdutil/string.cpp diff --git a/excuter/cpp-common/src/stdutil/string.hpp b/executor/cpp-common/src/stdutil/string.hpp similarity index 100% rename from excuter/cpp-common/src/stdutil/string.hpp rename to executor/cpp-common/src/stdutil/string.hpp diff --git a/excuter/cpp-common/src/stdutil/time.hpp b/executor/cpp-common/src/stdutil/time.hpp similarity index 100% rename from excuter/cpp-common/src/stdutil/time.hpp rename to executor/cpp-common/src/stdutil/time.hpp diff --git a/excuter/cpp-common/src/stdutil/vector.hpp b/executor/cpp-common/src/stdutil/vector.hpp similarity index 100% rename from excuter/cpp-common/src/stdutil/vector.hpp rename to executor/cpp-common/src/stdutil/vector.hpp diff --git a/excuter/cpp-common/test/0_dtypes.cpp b/executor/cpp-common/test/0_dtypes.cpp similarity index 100% rename from excuter/cpp-common/test/0_dtypes.cpp rename to executor/cpp-common/test/0_dtypes.cpp diff --git a/excuter/cpp-common/test/1_tf.cpp b/executor/cpp-common/test/1_tf.cpp similarity index 100% rename from excuter/cpp-common/test/1_tf.cpp rename to executor/cpp-common/test/1_tf.cpp diff --git a/excuter/cpp-common/test/1_tfcheck.cpp b/executor/cpp-common/test/1_tfcheck.cpp similarity index 100% rename from excuter/cpp-common/test/1_tfcheck.cpp rename to executor/cpp-common/test/1_tfcheck.cpp diff --git a/excuter/cpp-common/test/2_saveload.cpp b/executor/cpp-common/test/2_saveload.cpp similarity index 100% rename from excuter/cpp-common/test/2_saveload.cpp rename to executor/cpp-common/test/2_saveload.cpp diff --git a/excuter/cpp-common/test/CMakeLists.txt b/executor/cpp-common/test/CMakeLists.txt similarity index 100% rename from excuter/cpp-common/test/CMakeLists.txt rename to executor/cpp-common/test/CMakeLists.txt diff --git a/excuter/op-mem-cuda/.gitignore b/executor/op-mem-cuda/.gitignore similarity index 100% rename from excuter/op-mem-cuda/.gitignore rename to executor/op-mem-cuda/.gitignore diff --git a/excuter/op-mem-cuda/CMakeLists.txt b/executor/op-mem-cuda/CMakeLists.txt similarity index 100% rename from excuter/op-mem-cuda/CMakeLists.txt rename to executor/op-mem-cuda/CMakeLists.txt diff --git a/excuter/op-mem-cuda/Dockerfile b/executor/op-mem-cuda/Dockerfile similarity index 100% rename from excuter/op-mem-cuda/Dockerfile rename to executor/op-mem-cuda/Dockerfile diff --git a/excuter/op-mem-cuda/README.md b/executor/op-mem-cuda/README.md similarity index 100% rename from excuter/op-mem-cuda/README.md rename to executor/op-mem-cuda/README.md diff --git a/excuter/op-mem-cuda/build.sh b/executor/op-mem-cuda/build.sh similarity index 100% rename from excuter/op-mem-cuda/build.sh rename to executor/op-mem-cuda/build.sh diff --git a/excuter/op-mem-cuda/doc/00_quickstart.md b/executor/op-mem-cuda/doc/00_quickstart.md similarity index 100% rename from excuter/op-mem-cuda/doc/00_quickstart.md rename to executor/op-mem-cuda/doc/00_quickstart.md diff --git a/excuter/op-mem-cuda/doc/01_layout.md b/executor/op-mem-cuda/doc/01_layout.md similarity index 100% rename from excuter/op-mem-cuda/doc/01_layout.md rename to executor/op-mem-cuda/doc/01_layout.md diff --git a/excuter/op-mem-cuda/doc/02_layout_algebra.md b/executor/op-mem-cuda/doc/02_layout_algebra.md similarity index 100% rename from excuter/op-mem-cuda/doc/02_layout_algebra.md rename to executor/op-mem-cuda/doc/02_layout_algebra.md diff --git a/excuter/op-mem-cuda/doc/03_tensor.md b/executor/op-mem-cuda/doc/03_tensor.md similarity index 100% rename from excuter/op-mem-cuda/doc/03_tensor.md rename to executor/op-mem-cuda/doc/03_tensor.md diff --git a/excuter/op-mem-cuda/doc/04_algorithms.md b/executor/op-mem-cuda/doc/04_algorithms.md similarity index 100% rename from excuter/op-mem-cuda/doc/04_algorithms.md rename to executor/op-mem-cuda/doc/04_algorithms.md diff --git a/excuter/op-mem-cuda/doc/0t_mma_atom.md b/executor/op-mem-cuda/doc/0t_mma_atom.md similarity index 100% rename from excuter/op-mem-cuda/doc/0t_mma_atom.md rename to executor/op-mem-cuda/doc/0t_mma_atom.md diff --git a/excuter/op-mem-cuda/doc/0x_gemm_tutorial.md b/executor/op-mem-cuda/doc/0x_gemm_tutorial.md similarity index 100% rename from excuter/op-mem-cuda/doc/0x_gemm_tutorial.md rename to executor/op-mem-cuda/doc/0x_gemm_tutorial.md diff --git a/excuter/op-mem-cuda/doc/0y_predication.md b/executor/op-mem-cuda/doc/0y_predication.md similarity index 100% rename from excuter/op-mem-cuda/doc/0y_predication.md rename to executor/op-mem-cuda/doc/0y_predication.md diff --git a/excuter/op-mem-cuda/doc/0z_tma_tensors.md b/executor/op-mem-cuda/doc/0z_tma_tensors.md similarity index 100% rename from excuter/op-mem-cuda/doc/0z_tma_tensors.md rename to executor/op-mem-cuda/doc/0z_tma_tensors.md diff --git a/excuter/op-mem-cuda/dockerbuild.sh b/executor/op-mem-cuda/dockerbuild.sh similarity index 100% rename from excuter/op-mem-cuda/dockerbuild.sh rename to executor/op-mem-cuda/dockerbuild.sh diff --git a/excuter/op-mem-cuda/log.md b/executor/op-mem-cuda/log.md similarity index 100% rename from excuter/op-mem-cuda/log.md rename to executor/op-mem-cuda/log.md diff --git a/excuter/op-mem-cuda/src/client/main.cpp b/executor/op-mem-cuda/src/client/main.cpp similarity index 100% rename from excuter/op-mem-cuda/src/client/main.cpp rename to executor/op-mem-cuda/src/client/main.cpp diff --git a/excuter/op-mem-cuda/src/client/tfs.cpp b/executor/op-mem-cuda/src/client/tfs.cpp similarity index 100% rename from excuter/op-mem-cuda/src/client/tfs.cpp rename to executor/op-mem-cuda/src/client/tfs.cpp diff --git a/excuter/op-mem-cuda/src/client/tfs.hpp b/executor/op-mem-cuda/src/client/tfs.hpp similarity index 100% rename from excuter/op-mem-cuda/src/client/tfs.hpp rename to executor/op-mem-cuda/src/client/tfs.hpp diff --git a/excuter/op-mem-cuda/src/deepx/dtype_cuda.hpp b/executor/op-mem-cuda/src/deepx/dtype_cuda.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/dtype_cuda.hpp rename to executor/op-mem-cuda/src/deepx/dtype_cuda.hpp diff --git a/excuter/op-mem-cuda/src/deepx/mem/mem_cuda.hpp b/executor/op-mem-cuda/src/deepx/mem/mem_cuda.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/mem/mem_cuda.hpp rename to executor/op-mem-cuda/src/deepx/mem/mem_cuda.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/changeshape_miaobyte.cu b/executor/op-mem-cuda/src/deepx/tensorfunc/changeshape_miaobyte.cu similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/changeshape_miaobyte.cu rename to executor/op-mem-cuda/src/deepx/tensorfunc/changeshape_miaobyte.cu diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/changeshape_miaobyte.cuh b/executor/op-mem-cuda/src/deepx/tensorfunc/changeshape_miaobyte.cuh similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/changeshape_miaobyte.cuh rename to executor/op-mem-cuda/src/deepx/tensorfunc/changeshape_miaobyte.cuh diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/changeshape_miaobyte.hpp b/executor/op-mem-cuda/src/deepx/tensorfunc/changeshape_miaobyte.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/changeshape_miaobyte.hpp rename to executor/op-mem-cuda/src/deepx/tensorfunc/changeshape_miaobyte.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/cuda.hpp b/executor/op-mem-cuda/src/deepx/tensorfunc/cuda.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/cuda.hpp rename to executor/op-mem-cuda/src/deepx/tensorfunc/cuda.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/cuda_atomic.cuh b/executor/op-mem-cuda/src/deepx/tensorfunc/cuda_atomic.cuh similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/cuda_atomic.cuh rename to executor/op-mem-cuda/src/deepx/tensorfunc/cuda_atomic.cuh diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/cuda_math.cuh b/executor/op-mem-cuda/src/deepx/tensorfunc/cuda_math.cuh similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/cuda_math.cuh rename to executor/op-mem-cuda/src/deepx/tensorfunc/cuda_math.cuh diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_cublas_basic.hpp b/executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_cublas_basic.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_cublas_basic.hpp rename to executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_cublas_basic.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_basic.cu b/executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_basic.cu similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_basic.cu rename to executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_basic.cu diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_basic.cuh b/executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_basic.cuh similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_basic.cuh rename to executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_basic.cuh diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_basic.hpp b/executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_basic.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_basic.hpp rename to executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_basic.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_compare.cu b/executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_compare.cu similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_compare.cu rename to executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_compare.cu diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_compare.cuh b/executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_compare.cuh similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_compare.cuh rename to executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_compare.cuh diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_compare.hpp b/executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_compare.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_compare.hpp rename to executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_compare.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sin.cu b/executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sin.cu similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sin.cu rename to executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sin.cu diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sin.cuh b/executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sin.cuh similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sin.cuh rename to executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sin.cuh diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sin.hpp b/executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sin.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sin.hpp rename to executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sin.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sqrt.cu b/executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sqrt.cu similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sqrt.cu rename to executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sqrt.cu diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sqrt.cuh b/executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sqrt.cuh similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sqrt.cuh rename to executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sqrt.cuh diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sqrt.hpp b/executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sqrt.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sqrt.hpp rename to executor/op-mem-cuda/src/deepx/tensorfunc/elementwise_miaobyte_sqrt.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.cu b/executor/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.cu similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.cu rename to executor/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.cu diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.cuh b/executor/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.cuh similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.cuh rename to executor/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.cuh diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.hpp b/executor/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.hpp rename to executor/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/io_miaobyte.hpp b/executor/op-mem-cuda/src/deepx/tensorfunc/io_miaobyte.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/io_miaobyte.hpp rename to executor/op-mem-cuda/src/deepx/tensorfunc/io_miaobyte.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/matmul_cublas.hpp b/executor/op-mem-cuda/src/deepx/tensorfunc/matmul_cublas.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/matmul_cublas.hpp rename to executor/op-mem-cuda/src/deepx/tensorfunc/matmul_cublas.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/new_mempool.hpp b/executor/op-mem-cuda/src/deepx/tensorfunc/new_mempool.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/new_mempool.hpp rename to executor/op-mem-cuda/src/deepx/tensorfunc/new_mempool.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/reduce_miaobyte.cu b/executor/op-mem-cuda/src/deepx/tensorfunc/reduce_miaobyte.cu similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/reduce_miaobyte.cu rename to executor/op-mem-cuda/src/deepx/tensorfunc/reduce_miaobyte.cu diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/reduce_miaobyte.cuh b/executor/op-mem-cuda/src/deepx/tensorfunc/reduce_miaobyte.cuh similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/reduce_miaobyte.cuh rename to executor/op-mem-cuda/src/deepx/tensorfunc/reduce_miaobyte.cuh diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/reduce_miaobyte.hpp b/executor/op-mem-cuda/src/deepx/tensorfunc/reduce_miaobyte.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/reduce_miaobyte.hpp rename to executor/op-mem-cuda/src/deepx/tensorfunc/reduce_miaobyte.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/tensor_cuda.cuh b/executor/op-mem-cuda/src/deepx/tensorfunc/tensor_cuda.cuh similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/tensor_cuda.cuh rename to executor/op-mem-cuda/src/deepx/tensorfunc/tensor_cuda.cuh diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/tensorlife_miaobyte.hpp b/executor/op-mem-cuda/src/deepx/tensorfunc/tensorlife_miaobyte.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/tensorlife_miaobyte.hpp rename to executor/op-mem-cuda/src/deepx/tensorfunc/tensorlife_miaobyte.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/vector_cuda.cuh b/executor/op-mem-cuda/src/deepx/tensorfunc/vector_cuda.cuh similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tensorfunc/vector_cuda.cuh rename to executor/op-mem-cuda/src/deepx/tensorfunc/vector_cuda.cuh diff --git a/excuter/op-mem-cuda/src/deepx/tf/arg.hpp b/executor/op-mem-cuda/src/deepx/tf/arg.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tf/arg.hpp rename to executor/op-mem-cuda/src/deepx/tf/arg.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tf/changeshape.hpp b/executor/op-mem-cuda/src/deepx/tf/changeshape.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tf/changeshape.hpp rename to executor/op-mem-cuda/src/deepx/tf/changeshape.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tf/elementwise_basic.hpp b/executor/op-mem-cuda/src/deepx/tf/elementwise_basic.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tf/elementwise_basic.hpp rename to executor/op-mem-cuda/src/deepx/tf/elementwise_basic.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tf/elementwise_compare.hpp b/executor/op-mem-cuda/src/deepx/tf/elementwise_compare.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tf/elementwise_compare.hpp rename to executor/op-mem-cuda/src/deepx/tf/elementwise_compare.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tf/elementwise_sin.hpp b/executor/op-mem-cuda/src/deepx/tf/elementwise_sin.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tf/elementwise_sin.hpp rename to executor/op-mem-cuda/src/deepx/tf/elementwise_sin.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tf/elementwise_sqrt.hpp b/executor/op-mem-cuda/src/deepx/tf/elementwise_sqrt.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tf/elementwise_sqrt.hpp rename to executor/op-mem-cuda/src/deepx/tf/elementwise_sqrt.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tf/init.hpp b/executor/op-mem-cuda/src/deepx/tf/init.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tf/init.hpp rename to executor/op-mem-cuda/src/deepx/tf/init.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tf/io.hpp b/executor/op-mem-cuda/src/deepx/tf/io.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tf/io.hpp rename to executor/op-mem-cuda/src/deepx/tf/io.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tf/matmul.hpp b/executor/op-mem-cuda/src/deepx/tf/matmul.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tf/matmul.hpp rename to executor/op-mem-cuda/src/deepx/tf/matmul.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tf/reduce.hpp b/executor/op-mem-cuda/src/deepx/tf/reduce.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tf/reduce.hpp rename to executor/op-mem-cuda/src/deepx/tf/reduce.hpp diff --git a/excuter/op-mem-cuda/src/deepx/tf/tensorlife.hpp b/executor/op-mem-cuda/src/deepx/tf/tensorlife.hpp similarity index 100% rename from excuter/op-mem-cuda/src/deepx/tf/tensorlife.hpp rename to executor/op-mem-cuda/src/deepx/tf/tensorlife.hpp diff --git a/excuter/op-mem-cuda/test/op/CMakeLists.txt b/executor/op-mem-cuda/test/op/CMakeLists.txt similarity index 100% rename from excuter/op-mem-cuda/test/op/CMakeLists.txt rename to executor/op-mem-cuda/test/op/CMakeLists.txt diff --git a/excuter/op-mem-cuda/test/tensorfunc/0_new.cpp b/executor/op-mem-cuda/test/tensorfunc/0_new.cpp similarity index 100% rename from excuter/op-mem-cuda/test/tensorfunc/0_new.cpp rename to executor/op-mem-cuda/test/tensorfunc/0_new.cpp diff --git a/excuter/op-mem-cuda/test/tensorfunc/1_cublas_add.cpp b/executor/op-mem-cuda/test/tensorfunc/1_cublas_add.cpp similarity index 100% rename from excuter/op-mem-cuda/test/tensorfunc/1_cublas_add.cpp rename to executor/op-mem-cuda/test/tensorfunc/1_cublas_add.cpp diff --git a/excuter/op-mem-cuda/test/tensorfunc/1_cublas_matmul.cpp b/executor/op-mem-cuda/test/tensorfunc/1_cublas_matmul.cpp similarity index 100% rename from excuter/op-mem-cuda/test/tensorfunc/1_cublas_matmul.cpp rename to executor/op-mem-cuda/test/tensorfunc/1_cublas_matmul.cpp diff --git a/excuter/op-mem-cuda/test/tensorfunc/2_changeshape.cpp b/executor/op-mem-cuda/test/tensorfunc/2_changeshape.cpp similarity index 100% rename from excuter/op-mem-cuda/test/tensorfunc/2_changeshape.cpp rename to executor/op-mem-cuda/test/tensorfunc/2_changeshape.cpp diff --git a/excuter/op-mem-cuda/test/tensorfunc/CMakeLists.txt b/executor/op-mem-cuda/test/tensorfunc/CMakeLists.txt similarity index 100% rename from excuter/op-mem-cuda/test/tensorfunc/CMakeLists.txt rename to executor/op-mem-cuda/test/tensorfunc/CMakeLists.txt diff --git a/excuter/op-mem-mps/swift.md b/executor/op-mem-mps/swift.md similarity index 100% rename from excuter/op-mem-mps/swift.md rename to executor/op-mem-mps/swift.md diff --git a/excuter/op-mem-ompsimd/.cursorignore b/executor/op-mem-ompsimd/.cursorignore similarity index 100% rename from excuter/op-mem-ompsimd/.cursorignore rename to executor/op-mem-ompsimd/.cursorignore diff --git a/excuter/op-mem-ompsimd/.cursorrules b/executor/op-mem-ompsimd/.cursorrules similarity index 100% rename from excuter/op-mem-ompsimd/.cursorrules rename to executor/op-mem-ompsimd/.cursorrules diff --git a/excuter/op-mem-ompsimd/.gitignore b/executor/op-mem-ompsimd/.gitignore similarity index 100% rename from excuter/op-mem-ompsimd/.gitignore rename to executor/op-mem-ompsimd/.gitignore diff --git a/excuter/op-mem-ompsimd/CMakeLists.txt b/executor/op-mem-ompsimd/CMakeLists.txt similarity index 100% rename from excuter/op-mem-ompsimd/CMakeLists.txt rename to executor/op-mem-ompsimd/CMakeLists.txt diff --git a/excuter/op-mem-ompsimd/Dockerfile b/executor/op-mem-ompsimd/Dockerfile similarity index 100% rename from excuter/op-mem-ompsimd/Dockerfile rename to executor/op-mem-ompsimd/Dockerfile diff --git a/excuter/op-mem-ompsimd/dockerbuild.sh b/executor/op-mem-ompsimd/dockerbuild.sh similarity index 100% rename from excuter/op-mem-ompsimd/dockerbuild.sh rename to executor/op-mem-ompsimd/dockerbuild.sh diff --git a/excuter/op-mem-ompsimd/log.md b/executor/op-mem-ompsimd/log.md similarity index 100% rename from excuter/op-mem-ompsimd/log.md rename to executor/op-mem-ompsimd/log.md diff --git a/excuter/op-mem-ompsimd/src/client/main.cpp b/executor/op-mem-ompsimd/src/client/main.cpp similarity index 100% rename from excuter/op-mem-ompsimd/src/client/main.cpp rename to executor/op-mem-ompsimd/src/client/main.cpp diff --git a/excuter/op-mem-ompsimd/src/client/tfs.cpp b/executor/op-mem-ompsimd/src/client/tfs.cpp similarity index 100% rename from excuter/op-mem-ompsimd/src/client/tfs.cpp rename to executor/op-mem-ompsimd/src/client/tfs.cpp diff --git a/excuter/op-mem-ompsimd/src/client/tfs.hpp b/executor/op-mem-ompsimd/src/client/tfs.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/client/tfs.hpp rename to executor/op-mem-ompsimd/src/client/tfs.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/dtype_ompsimd.hpp b/executor/op-mem-ompsimd/src/deepx/dtype_ompsimd.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/dtype_ompsimd.hpp rename to executor/op-mem-ompsimd/src/deepx/dtype_ompsimd.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/mem/mem_ompsimd.hpp b/executor/op-mem-ompsimd/src/deepx/mem/mem_ompsimd.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/mem/mem_ompsimd.hpp rename to executor/op-mem-ompsimd/src/deepx/mem/mem_ompsimd.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tensorfunc/changeshape_miaobyte.hpp b/executor/op-mem-ompsimd/src/deepx/tensorfunc/changeshape_miaobyte.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tensorfunc/changeshape_miaobyte.hpp rename to executor/op-mem-ompsimd/src/deepx/tensorfunc/changeshape_miaobyte.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tensorfunc/elementwise_cblas.hpp b/executor/op-mem-ompsimd/src/deepx/tensorfunc/elementwise_cblas.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tensorfunc/elementwise_cblas.hpp rename to executor/op-mem-ompsimd/src/deepx/tensorfunc/elementwise_cblas.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tensorfunc/elementwise_miaobyte.hpp b/executor/op-mem-ompsimd/src/deepx/tensorfunc/elementwise_miaobyte.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tensorfunc/elementwise_miaobyte.hpp rename to executor/op-mem-ompsimd/src/deepx/tensorfunc/elementwise_miaobyte.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tensorfunc/equal.hpp b/executor/op-mem-ompsimd/src/deepx/tensorfunc/equal.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tensorfunc/equal.hpp rename to executor/op-mem-ompsimd/src/deepx/tensorfunc/equal.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tensorfunc/highway.hpp b/executor/op-mem-ompsimd/src/deepx/tensorfunc/highway.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tensorfunc/highway.hpp rename to executor/op-mem-ompsimd/src/deepx/tensorfunc/highway.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tensorfunc/init_miaobyte.hpp b/executor/op-mem-ompsimd/src/deepx/tensorfunc/init_miaobyte.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tensorfunc/init_miaobyte.hpp rename to executor/op-mem-ompsimd/src/deepx/tensorfunc/init_miaobyte.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tensorfunc/io_miaobyte.hpp b/executor/op-mem-ompsimd/src/deepx/tensorfunc/io_miaobyte.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tensorfunc/io_miaobyte.hpp rename to executor/op-mem-ompsimd/src/deepx/tensorfunc/io_miaobyte.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tensorfunc/matmul_cblas.hpp b/executor/op-mem-ompsimd/src/deepx/tensorfunc/matmul_cblas.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tensorfunc/matmul_cblas.hpp rename to executor/op-mem-ompsimd/src/deepx/tensorfunc/matmul_cblas.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tensorfunc/matmul_miaobyte.hpp b/executor/op-mem-ompsimd/src/deepx/tensorfunc/matmul_miaobyte.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tensorfunc/matmul_miaobyte.hpp rename to executor/op-mem-ompsimd/src/deepx/tensorfunc/matmul_miaobyte.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tensorfunc/new_mempool.hpp b/executor/op-mem-ompsimd/src/deepx/tensorfunc/new_mempool.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tensorfunc/new_mempool.hpp rename to executor/op-mem-ompsimd/src/deepx/tensorfunc/new_mempool.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tensorfunc/reduce_miaobyte.hpp b/executor/op-mem-ompsimd/src/deepx/tensorfunc/reduce_miaobyte.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tensorfunc/reduce_miaobyte.hpp rename to executor/op-mem-ompsimd/src/deepx/tensorfunc/reduce_miaobyte.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tensorfunc/tensorlife_miaobyte.hpp b/executor/op-mem-ompsimd/src/deepx/tensorfunc/tensorlife_miaobyte.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tensorfunc/tensorlife_miaobyte.hpp rename to executor/op-mem-ompsimd/src/deepx/tensorfunc/tensorlife_miaobyte.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tf/arg.hpp b/executor/op-mem-ompsimd/src/deepx/tf/arg.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tf/arg.hpp rename to executor/op-mem-ompsimd/src/deepx/tf/arg.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tf/changeshape.hpp b/executor/op-mem-ompsimd/src/deepx/tf/changeshape.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tf/changeshape.hpp rename to executor/op-mem-ompsimd/src/deepx/tf/changeshape.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tf/elementwise.hpp b/executor/op-mem-ompsimd/src/deepx/tf/elementwise.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tf/elementwise.hpp rename to executor/op-mem-ompsimd/src/deepx/tf/elementwise.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tf/init.hpp b/executor/op-mem-ompsimd/src/deepx/tf/init.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tf/init.hpp rename to executor/op-mem-ompsimd/src/deepx/tf/init.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tf/io.hpp b/executor/op-mem-ompsimd/src/deepx/tf/io.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tf/io.hpp rename to executor/op-mem-ompsimd/src/deepx/tf/io.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tf/matmul.hpp b/executor/op-mem-ompsimd/src/deepx/tf/matmul.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tf/matmul.hpp rename to executor/op-mem-ompsimd/src/deepx/tf/matmul.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tf/reduce.hpp b/executor/op-mem-ompsimd/src/deepx/tf/reduce.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tf/reduce.hpp rename to executor/op-mem-ompsimd/src/deepx/tf/reduce.hpp diff --git a/excuter/op-mem-ompsimd/src/deepx/tf/tensorlife.hpp b/executor/op-mem-ompsimd/src/deepx/tf/tensorlife.hpp similarity index 100% rename from excuter/op-mem-ompsimd/src/deepx/tf/tensorlife.hpp rename to executor/op-mem-ompsimd/src/deepx/tf/tensorlife.hpp diff --git a/excuter/op-mem-ompsimd/test/op/1_mem.cpp b/executor/op-mem-ompsimd/test/op/1_mem.cpp similarity index 100% rename from excuter/op-mem-ompsimd/test/op/1_mem.cpp rename to executor/op-mem-ompsimd/test/op/1_mem.cpp diff --git a/excuter/op-mem-ompsimd/test/op/CMakeLists.txt b/executor/op-mem-ompsimd/test/op/CMakeLists.txt similarity index 100% rename from excuter/op-mem-ompsimd/test/op/CMakeLists.txt rename to executor/op-mem-ompsimd/test/op/CMakeLists.txt diff --git a/excuter/op-mem-ompsimd/test/tensorfunc/1_shape.cpp b/executor/op-mem-ompsimd/test/tensorfunc/1_shape.cpp similarity index 100% rename from excuter/op-mem-ompsimd/test/tensorfunc/1_shape.cpp rename to executor/op-mem-ompsimd/test/tensorfunc/1_shape.cpp diff --git a/excuter/op-mem-ompsimd/test/tensorfunc/2_shape_combintion.cpp b/executor/op-mem-ompsimd/test/tensorfunc/2_shape_combintion.cpp similarity index 100% rename from excuter/op-mem-ompsimd/test/tensorfunc/2_shape_combintion.cpp rename to executor/op-mem-ompsimd/test/tensorfunc/2_shape_combintion.cpp diff --git a/excuter/op-mem-ompsimd/test/tensorfunc/2_tensor_equal.cpp b/executor/op-mem-ompsimd/test/tensorfunc/2_tensor_equal.cpp similarity index 100% rename from excuter/op-mem-ompsimd/test/tensorfunc/2_tensor_equal.cpp rename to executor/op-mem-ompsimd/test/tensorfunc/2_tensor_equal.cpp diff --git a/excuter/op-mem-ompsimd/test/tensorfunc/2_tensor_new.cpp b/executor/op-mem-ompsimd/test/tensorfunc/2_tensor_new.cpp similarity index 100% rename from excuter/op-mem-ompsimd/test/tensorfunc/2_tensor_new.cpp rename to executor/op-mem-ompsimd/test/tensorfunc/2_tensor_new.cpp diff --git a/excuter/op-mem-ompsimd/test/tensorfunc/2_tensor_range.cpp b/executor/op-mem-ompsimd/test/tensorfunc/2_tensor_range.cpp similarity index 100% rename from excuter/op-mem-ompsimd/test/tensorfunc/2_tensor_range.cpp rename to executor/op-mem-ompsimd/test/tensorfunc/2_tensor_range.cpp diff --git a/excuter/op-mem-ompsimd/test/tensorfunc/2_tensor_range.py b/executor/op-mem-ompsimd/test/tensorfunc/2_tensor_range.py similarity index 100% rename from excuter/op-mem-ompsimd/test/tensorfunc/2_tensor_range.py rename to executor/op-mem-ompsimd/test/tensorfunc/2_tensor_range.py diff --git a/excuter/op-mem-ompsimd/test/tensorfunc/3_tensor_print.cpp b/executor/op-mem-ompsimd/test/tensorfunc/3_tensor_print.cpp similarity index 100% rename from excuter/op-mem-ompsimd/test/tensorfunc/3_tensor_print.cpp rename to executor/op-mem-ompsimd/test/tensorfunc/3_tensor_print.cpp diff --git a/excuter/op-mem-ompsimd/test/tensorfunc/4_tensor_add.cpp b/executor/op-mem-ompsimd/test/tensorfunc/4_tensor_add.cpp similarity index 100% rename from excuter/op-mem-ompsimd/test/tensorfunc/4_tensor_add.cpp rename to executor/op-mem-ompsimd/test/tensorfunc/4_tensor_add.cpp diff --git a/excuter/op-mem-ompsimd/test/tensorfunc/4_tensor_matmul.cpp b/executor/op-mem-ompsimd/test/tensorfunc/4_tensor_matmul.cpp similarity index 100% rename from excuter/op-mem-ompsimd/test/tensorfunc/4_tensor_matmul.cpp rename to executor/op-mem-ompsimd/test/tensorfunc/4_tensor_matmul.cpp diff --git a/excuter/op-mem-ompsimd/test/tensorfunc/4_tensor_max.cpp b/executor/op-mem-ompsimd/test/tensorfunc/4_tensor_max.cpp similarity index 100% rename from excuter/op-mem-ompsimd/test/tensorfunc/4_tensor_max.cpp rename to executor/op-mem-ompsimd/test/tensorfunc/4_tensor_max.cpp diff --git a/excuter/op-mem-ompsimd/test/tensorfunc/4_tensor_mul.cpp b/executor/op-mem-ompsimd/test/tensorfunc/4_tensor_mul.cpp similarity index 100% rename from excuter/op-mem-ompsimd/test/tensorfunc/4_tensor_mul.cpp rename to executor/op-mem-ompsimd/test/tensorfunc/4_tensor_mul.cpp diff --git a/excuter/op-mem-ompsimd/test/tensorfunc/4_tensor_sub.cpp b/executor/op-mem-ompsimd/test/tensorfunc/4_tensor_sub.cpp similarity index 100% rename from excuter/op-mem-ompsimd/test/tensorfunc/4_tensor_sub.cpp rename to executor/op-mem-ompsimd/test/tensorfunc/4_tensor_sub.cpp diff --git a/excuter/op-mem-ompsimd/test/tensorfunc/5_tensor_sum.cpp b/executor/op-mem-ompsimd/test/tensorfunc/5_tensor_sum.cpp similarity index 100% rename from excuter/op-mem-ompsimd/test/tensorfunc/5_tensor_sum.cpp rename to executor/op-mem-ompsimd/test/tensorfunc/5_tensor_sum.cpp diff --git a/excuter/op-mem-ompsimd/test/tensorfunc/6_tensor_broadcast.cpp b/executor/op-mem-ompsimd/test/tensorfunc/6_tensor_broadcast.cpp similarity index 100% rename from excuter/op-mem-ompsimd/test/tensorfunc/6_tensor_broadcast.cpp rename to executor/op-mem-ompsimd/test/tensorfunc/6_tensor_broadcast.cpp diff --git a/excuter/op-mem-ompsimd/test/tensorfunc/7_tensor_transpose.cpp b/executor/op-mem-ompsimd/test/tensorfunc/7_tensor_transpose.cpp similarity index 100% rename from excuter/op-mem-ompsimd/test/tensorfunc/7_tensor_transpose.cpp rename to executor/op-mem-ompsimd/test/tensorfunc/7_tensor_transpose.cpp diff --git a/excuter/op-mem-ompsimd/test/tensorfunc/8_tensor_concat.cpp b/executor/op-mem-ompsimd/test/tensorfunc/8_tensor_concat.cpp similarity index 100% rename from excuter/op-mem-ompsimd/test/tensorfunc/8_tensor_concat.cpp rename to executor/op-mem-ompsimd/test/tensorfunc/8_tensor_concat.cpp diff --git a/excuter/op-mem-ompsimd/test/tensorfunc/CMakeLists.txt b/executor/op-mem-ompsimd/test/tensorfunc/CMakeLists.txt similarity index 100% rename from excuter/op-mem-ompsimd/test/tensorfunc/CMakeLists.txt rename to executor/op-mem-ompsimd/test/tensorfunc/CMakeLists.txt diff --git a/excuter/op-mem-ompsimd/test/tensorfunc/tensorutil.hpp b/executor/op-mem-ompsimd/test/tensorfunc/tensorutil.hpp similarity index 100% rename from excuter/op-mem-ompsimd/test/tensorfunc/tensorutil.hpp rename to executor/op-mem-ompsimd/test/tensorfunc/tensorutil.hpp From 6b5bd169af85b09eba7dc449dde846b5bae7d902 Mon Sep 17 00:00:00 2001 From: LJXjean <1372682462@qq.com> Date: Sat, 5 Jul 2025 15:56:10 +0200 Subject: [PATCH 2/6] all "excuter" turned into "executer" --- .cursorrules | 2 +- .../{excuter.md => executor.md} | 4 +- ...r-cppcommon.yml => executor-cppcommon.yml} | 8 +- ...cuda-linux.yml => executor-cuda-linux.yml} | 6 +- ...d-linux.yml => executor-ompsimd-linux.yml} | 12 +- CONTRIBUTING.md | 2 +- README.md | 10 +- all_doc | 7 + all_excuter.txt | 297 ++++++++++++++++++ docs/deepxIR/ir.md | 14 +- docs/design.md | 4 +- docs/executor/deepx.op.drawio | 6 +- docs/executor/deepx.op.drawio.svg | 2 +- docs/executor/{excuter.md => executor.md} | 10 +- docs/executor/mix_precision.md | 2 +- docs/executor/op-mem-cuda/list.md | 2 +- docs/executor/op-mem-ompsimd/contribute.md | 2 +- docs/executor/op-mem-ompsimd/list.md | 2 +- docs/executor/op-mem-ompsimd/range.md | 6 +- docs/front/deepx.op.drawio.svg | 2 +- docs/index.rst | 8 +- docs/language.md | 6 +- docs/scheduler/scheduler.md | 6 +- .../cpp-common/src/deepx/tf/tffactory.cpp | 6 +- .../cpp-common/src/deepx/tf/tffactory.hpp | 2 +- executor/op-mem-cuda/CMakeLists.txt | 2 +- executor/op-mem-cuda/Dockerfile | 2 +- executor/op-mem-cuda/README.md | 2 +- executor/op-mem-cuda/src/client/main.cpp | 2 +- executor/op-mem-ompsimd/.cursorrules | 2 +- executor/op-mem-ompsimd/CMakeLists.txt | 2 +- executor/op-mem-ompsimd/Dockerfile | 2 +- executor/op-mem-ompsimd/src/client/main.cpp | 2 +- front/go/README.md | 2 +- front/py/.cursorrules | 6 +- front/py/deepx/README.md | 2 +- rename.py | 18 ++ 37 files changed, 397 insertions(+), 75 deletions(-) rename .github/ISSUE_TEMPLATE/{excuter.md => executor.md} (89%) rename .github/workflows/{excuter-cppcommon.yml => executor-cppcommon.yml} (91%) rename .github/workflows/{excuter-cuda-linux.yml => executor-cuda-linux.yml} (96%) rename .github/workflows/{excuter-ompsimd-linux.yml => executor-ompsimd-linux.yml} (91%) create mode 100644 all_doc create mode 100644 all_excuter.txt rename docs/executor/{excuter.md => executor.md} (81%) create mode 100644 rename.py diff --git a/.cursorrules b/.cursorrules index cf7f5ee0..a26d2166 100644 --- a/.cursorrules +++ b/.cursorrules @@ -6,7 +6,7 @@ Always respond in 中文 项目分为3部分 1. 前端。python库的接口风格参考pytorch，其他语言如go,java,c,rust等，后续设计完善。 2. 调度器，待设计 -3. 执行器，使用c++,cuda,metal,omp simd等,实现不同excuter的算子的前向和反向 +3. 执行器，使用c++,cuda,metal,omp simd等,实现不同executor的算子的前向和反向 关于概念 deepx.Tensor仅仅就是一个tensor，不像pytorch的tensor，一个tensor其实包含了自身和梯度2个tensor的数据 diff --git a/.github/ISSUE_TEMPLATE/excuter.md b/.github/ISSUE_TEMPLATE/executor.md similarity index 89% rename from .github/ISSUE_TEMPLATE/excuter.md rename to .github/ISSUE_TEMPLATE/executor.md index a3310831..a32ee694 100644 --- a/.github/ISSUE_TEMPLATE/excuter.md +++ b/.github/ISSUE_TEMPLATE/executor.md @@ -1,8 +1,8 @@ --- name: 执行引擎 about:按照给定计算图，负责存储、计算、网络传输的执行 -title: '[excuter] ' -labels: excuter, +title: '[executor] ' +labels: executor, assignees: '' --- diff --git a/.github/workflows/excuter-cppcommon.yml b/.github/workflows/executor-cppcommon.yml similarity index 91% rename from .github/workflows/excuter-cppcommon.yml rename to .github/workflows/executor-cppcommon.yml index 5eb86933..efa39092 100644 --- a/.github/workflows/excuter-cppcommon.yml +++ b/.github/workflows/executor-cppcommon.yml @@ -2,10 +2,10 @@ name: Excuter/cppcommon Build on: [push, pull_request] push: paths: - - 'excuter/cpp-common/**' + - 'executor/cpp-common/**' pull_request: paths: - - 'excuter/cpp-common/**' + - 'executor/cpp-common/**' env: HIGHWAY_VERSION: 1.2.0 @@ -48,7 +48,7 @@ jobs: uses: actions/cache@v3 with: path: | - excuter/cpp-common/build + executor/cpp-common/build ~/.ccache key: ${{ runner.os }}-build-${{ hashFiles('**/CMakeLists.txt') }} restore-keys: | @@ -57,7 +57,7 @@ jobs: # 构建 cpp-common 库 - name: Build Common Library run: | - cd excuter/cpp-common + cd executor/cpp-common mkdir -p build && cd build cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_CXX_COMPILER_LAUNCHER=ccache .. cmake --build . --config Release -j$(nproc) diff --git a/.github/workflows/excuter-cuda-linux.yml b/.github/workflows/executor-cuda-linux.yml similarity index 96% rename from .github/workflows/excuter-cuda-linux.yml rename to .github/workflows/executor-cuda-linux.yml index 70ce9050..4bf3359c 100644 --- a/.github/workflows/excuter-cuda-linux.yml +++ b/.github/workflows/executor-cuda-linux.yml @@ -2,10 +2,10 @@ name: Excuter/cuda-linux Build on: [push, pull_request] push: paths: - - 'excuter/op-mem-cuda/**' + - 'executor/op-mem-cuda/**' pull_request: paths: - - 'excuter/op-mem-cuda/**' + - 'executor/op-mem-cuda/**' env: CUDA_VERSION: "12.6.0" CUDA_MAJOR_VERSION: "12" @@ -62,7 +62,7 @@ jobs: cd /workspace && \ # 构建 common 库 - cd excuter/cpp-common && \ + cd executor/cpp-common && \ mkdir -p build && cd build && \ cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_CXX_COMPILER_LAUNCHER=ccache -GNinja .. && \ ninja && \ diff --git a/.github/workflows/excuter-ompsimd-linux.yml b/.github/workflows/executor-ompsimd-linux.yml similarity index 91% rename from .github/workflows/excuter-ompsimd-linux.yml rename to .github/workflows/executor-ompsimd-linux.yml index 1376018f..4b4d08b7 100644 --- a/.github/workflows/excuter-ompsimd-linux.yml +++ b/.github/workflows/executor-ompsimd-linux.yml @@ -3,10 +3,10 @@ on: [push, pull_request] on: [push, pull_request] push: paths: - - 'excuter/op-mem-ompsimd/**' + - 'executor/op-mem-ompsimd/**' pull_request: paths: - - 'excuter/op-mem-ompsimd/**' + - 'executor/op-mem-ompsimd/**' env: HIGHWAY_VERSION: 1.2.0 @@ -49,8 +49,8 @@ jobs: uses: actions/cache@v3 with: path: | - excuter/op-mem-ompsimd/build - excuter/cpp-common/build + executor/op-mem-ompsimd/build + executor/cpp-common/build ~/.ccache key: ${{ runner.os }}-build-${{ hashFiles('**/CMakeLists.txt') }} restore-keys: | @@ -85,7 +85,7 @@ jobs: # 构建 cpp-common 库 - name: Build Common Library run: | - cd excuter/cpp-common + cd executor/cpp-common mkdir -p build && cd build cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_CXX_COMPILER_LAUNCHER=ccache .. cmake --build . --config Release -j$(nproc) @@ -93,7 +93,7 @@ jobs: # 构建执行器 - name: CMake Build run: | - cd excuter/op-mem-ompsimd + cd executor/op-mem-ompsimd mkdir -p build && cd build cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_CXX_COMPILER_LAUNCHER=ccache .. cmake --build . --config Release -j$(nproc) \ No newline at end of file diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md index 88efd26d..939adf76 100644 --- a/CONTRIBUTING.md +++ b/CONTRIBUTING.md @@ -4,7 +4,7 @@ deepx框架的发展，主要包括五大类方向 + front: 新增模型、module、python类函数等 + 中间层：包括计算图优化器，插件系统(自动KVcache系统)，自动分布式化，栈tensor自动释放，自动Inplace化等操作 -+ 新增或修改excuter ++ 新增或修改executor + 增加或修改算子，进一步可以分为leaftensorfunc(不可分割的基础算子)，fusedtensorfunc（融合算子） + 文档丰富： + 运维自动化方向 diff --git a/README.md b/README.md index fa009c8f..a24a727f 100644 --- a/README.md +++ b/README.md @@ -15,7 +15,7 @@ deepx的分前中后端，分别是为前端表达侧，编译替换层，执行 python sdk提供接近pytorch的API 也容许其他语言的sdk接入， -+ IR通信调度。不同于pytorch或其他py+bind c++这种单一进程的栈上函数调度执行的方式。deepx各个程序（如front的python sdk，back的计算图编译器优化器、excuter如ompsimd）之间，通过IR实现网络通信调度，需要各自启动对应进程。 ++ IR通信调度。不同于pytorch或其他py+bind c++这种单一进程的栈上函数调度执行的方式。deepx各个程序（如front的python sdk，back的计算图编译器优化器、executor如ompsimd）之间，通过IR实现网络通信调度，需要各自启动对应进程。 | 维度 | PyTorch类框架 | DeepX | @@ -30,7 +30,7 @@ python sdk提供接近pytorch的API + 注册中心:收集当前已就绪的执行器的算子列表,收集算子时耗和空间占用信息 + 计算图编译器优化器:fusion算子，计算图节点消除,自动生成tensor拆分并行的计算子图并替代原节点 + 执行调度器：数据并行，流水线并行(前向反向并行)，模型并行。 -+ front生成基础IR，编译器负责进行fusion成excuter注册的高级算子。 ++ front生成基础IR，编译器负责进行fusion成executor注册的高级算子。 ### 执行层 @@ -44,16 +44,16 @@ Op{args(args_grad),returns(returns_grad)|func run} Op需要实现run方法 -关于excuter，只要能按deepxIR序列执行，并返回结果，就可以接入deepx分布式调度框架，因此，从硬件、指令、加速库、高级框架包括训练、推理引擎，都可以稍作修改，就接入deepx体系。 +关于executor，只要能按deepxIR序列执行，并返回结果，就可以接入deepx分布式调度框架，因此，从硬件、指令、加速库、高级框架包括训练、推理引擎，都可以稍作修改，就接入deepx体系。 当前的 #### 默认执行器 -+ cpu执行器,已实现ompsimd。其支持的算子列表[ompsimd](doc/excuter/op-mem-ompsimd/list.md) ++ cpu执行器,已实现ompsimd。其支持的算子列表[ompsimd](doc/executor/op-mem-ompsimd/list.md) #### GPU执行器 -+ cuda执行器，其支持的算子列表[cuda](doc/excuter/op-mem-cuda/list.md) ++ cuda执行器，其支持的算子列表[cuda](doc/executor/op-mem-cuda/list.md) 欢迎大家提交cuda代码 diff --git a/all_doc b/all_doc new file mode 100644 index 00000000..83dc5613 --- /dev/null +++ b/all_doc @@ -0,0 +1,7 @@ +README.md:53:+ cpu执行器,已实现ompsimd。其支持的算子列表[ompsimd](doc/executor/op-mem-ompsimd/list.md) +README.md:56:+ cuda执行器，其支持的算子列表[cuda](doc/executor/op-mem-cuda/list.md) +docs/README.md:56:![架构图](./doc/deepx.jpg) +docs/front/deepxpy.drawio.svg:4:
autograd
autograd
functional

elementwise:add,sub..
matmul:
reduce:sum,min,max

functional...
tensor

shape
dtype
tensor...
OpNode
OpNode
DataNode
type:var,vector,tensor
data:
DataNode...
Node
Node
Graph
- nodes
Graph...
deepx-front-py
deepx-front-py
ControlflowNode
ControlflowNode
nn
nn
deepxIR
deepxIRText is not SVG - cannot display +docs/language.md:38:https://github.com/array2d/deepx/blob/main/doc/executor/op-mem-cuda/list.md +executor/op-mem-cuda/src/client/main.cpp:32: string docdir = "../../../doc/executor/op-mem-cuda/"; +executor/op-mem-ompsimd/src/client/main.cpp:33: string docdir = "../../../doc/executor/op-mem-ompsimd/"; diff --git a/all_excuter.txt b/all_excuter.txt new file mode 100644 index 00000000..341f230f --- /dev/null +++ b/all_excuter.txt @@ -0,0 +1,297 @@ +./README.md:18:+ IR通信调度。不同于pytorch或其他py+bind c++这种单一进程的栈上函数调度执行的方式。deepx各个程序（如front的python sdk，back的计算图编译器优化器、executor如ompsimd）之间，通过IR实现网络通信调度，需要各自启动对应进程。 +./README.md:33:+ front生成基础IR，编译器负责进行fusion成executor注册的高级算子。 +./README.md:47:关于executor，只要能按deepxIR序列执行，并返回结果，就可以接入deepx分布式调度框架，因此，从硬件、指令、加速库、高级框架包括训练、推理引擎，都可以稍作修改，就接入deepx体系。 +./README.md:53:+ cpu执行器,已实现ompsimd。其支持的算子列表[ompsimd](doc/executor/op-mem-ompsimd/list.md) +./README.md:56:+ cuda执行器，其支持的算子列表[cuda](doc/executor/op-mem-cuda/list.md) +./docs/language.md:1:## c++:计算执行器(executor) +./docs/language.md:36:executor所执行的内容，就是deepxir的序列或deepxir计算图 +./docs/language.md:38:https://github.com/array2d/deepx/blob/main/doc/executor/op-mem-cuda/list.md +./docs/scheduler/scheduler.md:3:DeepX框架的scheduler，是front和executor之间的桥梁。 +./docs/scheduler/scheduler.md:5:front只负责搭建抽象计算图，executor负责执行算子，而scheduler负责将抽象计算图转换为执行计算图，并发送给executor。 +./docs/scheduler/scheduler.md:9:算子注册器，接收executor的算子及精度列表。 +./docs/executor/mix_precision.md:11:## 3. 关于executor的mix precision的实现 +./docs/executor/op-mem-cuda/list.md:3:本页面由 `executor/op-mem-cuda 生成，请勿手动修改 +./docs/executor/deepx.op.drawio.svg:4:
executor 
ompsimd
tensorfunc<T>
tensorfunc 特化
TF
cpp-common
executor 
cuda
 +./docs/executor/deepx.op.drawio:28: +./docs/executor/deepx.op.drawio:135: +./docs/executor/executor.md:1:## 如何给executor添加一个新算子 +./docs/executor/executor.md:17:Op是executor的算子，是executor的执行单元 +./docs/executor/executor.md:34:cd deepx/executor/op-mem-ompsimd +./docs/executor/executor.md:46:cd deepx/executor/op-mem-cuda +./docs/executor/executor.md:63:1.先启动executor可执行文件， 位于executor/op-mem-{cuda/ompsimd}/build，可执行文件名同executor名 +./docs/design.md:18:executor执行deepxIR的规则 +./docs/design.md:20:+ executor执行deepxIR时，不得修改args中的tensor +./docs/front/deepx.op.drawio.svg:4:
executor 
cuda
executor 
cpu
executor 
cuda
 T1 ## 函数定义(funcdef) -函数定义由excuter层负责注册实现,用于声明操作的参数和返回值类型。excuter通过注册funcdef来声明其支持的tensorfunc。 +函数定义由executor层负责注册实现,用于声明操作的参数和返回值类型。executor通过注册funcdef来声明其支持的tensorfunc。 因此需要设置参数、返回值的详细类型约束 @@ -50,7 +50,7 @@ matmul(A,B)->C //id=1 created_at=123456789 sent_at=123456790 对于tensorfunc的类型系统，我们只关心与tensor相关的类型系统 -参考 excuter/common/src/deepx/dtype.hpp +参考 executor/common/src/deepx/dtype.hpp ``` { @@ -79,7 +79,7 @@ matmul(A,B)->C //id=1 created_at=123456789 sent_at=123456790 ## funcdef -excuter 负责定义其支持的tensorfunc +executor 负责定义其支持的tensorfunc 1. 矩阵乘法: ``` @@ -89,7 +89,7 @@ matmul(Tensor A, Tensor B) -> Tensor C // rtf(remote tensor func)解析器会自动解析参数和返回值的列表 -// excuter会从mem获取A，B，C这3个tensor，并执行matmul操作 +// executor会从mem获取A，B，C这3个tensor，并执行matmul操作 ``` 2. 张量求和: @@ -100,9 +100,9 @@ sum(Tensor input, vector dims,var keepdim) -> Tensor outp # funccall sum(T1,[0 1],true) -> T2 // rtf(remote tensor func)解析器会自动解析参数和返回值的列表 -// 其中[0 1]会被解析为vector，便于excuter执行时使用 -// true会被解析为var keepdim，便于excuter执行时使用 -// excuter会从mem获取T1，T2这2个tensor，并执行sum操作 +// 其中[0 1]会被解析为vector，便于executor执行时使用 +// true会被解析为var keepdim，便于executor执行时使用 +// executor会从mem获取T1，T2这2个tensor，并执行sum操作 ``` 3. 创建新张量: diff --git a/docs/design.md b/docs/design.md index 0c93fe33..a333fdce 100644 --- a/docs/design.md +++ b/docs/design.md @@ -15,9 +15,9 @@ deepIR{ } ``` -excuter执行deepxIR的规则 +executor执行deepxIR的规则 -+ excuter执行deepxIR时，不得修改args中的tensor ++ executor执行deepxIR时，不得修改args中的tensor + 但deepIR不限制args和returns中的Param同名，这样可以实现类似inplace的操作 diff --git a/docs/executor/deepx.op.drawio b/docs/executor/deepx.op.drawio index 86117f1f..b2d240be 100644 --- a/docs/executor/deepx.op.drawio +++ b/docs/executor/deepx.op.drawio @@ -4,7 +4,7 @@ - + @@ -25,7 +25,7 @@ - + @@ -132,7 +132,7 @@ - + diff --git a/docs/executor/deepx.op.drawio.svg b/docs/executor/deepx.op.drawio.svg index f6fe1fd0..8ba6ed8a 100644 --- a/docs/executor/deepx.op.drawio.svg +++ b/docs/executor/deepx.op.drawio.svg @@ -1,4 +1,4 @@ -
excuter 
ompsimd
tensorfunc<T>
tensorfunc 特化
TF
cpp-common
excuter 
cuda
TFfactory
List
init.hpp
elementwise.hpp
matmul.hpp

io.hpp

reduce.hpp

changeshape.hpp
authora
init_authora.hpp
elementwise_authora.hpp
matmul_authora.hpp

io_authora.hpp

reduce_authora.hpp

changeshape_authora.hpp
authorb
matmul_authora.hpp

io_authora.hpp

reduce_authora.hpp

changeshape_authora.hpp
TF
+ name
+ args
+ returns
template author 特化
%3CmxGraphModel%3E%3Croot%3E%3CmxCell%20id%3D%220%22%2F%3E%3CmxCell%20id%3D%221%22%20parent%3D%220%22%2F%3E%3CmxCell%20id%3D%222%22%20style%3D%22edgeStyle%3DorthogonalEdgeStyle%3Brounded%3D0%3BorthogonalLoop%3D1%3BjettySize%3Dauto%3Bhtml%3D1%3BentryX%3D0%3BentryY%3D0.123%3BentryDx%3D0%3BentryDy%3D0%3BentryPerimeter%3D0%3BexitX%3D1.006%3BexitY%3D0.145%3BexitDx%3D0%3BexitDy%3D0%3BexitPerimeter%3D0%3B%22%20edge%3D%221%22%20parent%3D%221%22%3E%3CmxGeometry%20relative%3D%221%22%20as%3D%22geometry%22%3E%3CmxPoint%20x%3D%22-332%22%20y%3D%22283%22%20as%3D%22sourcePoint%22%2F%3E%3CmxPoint%20x%3D%22-150%22%20y%3D%22280%22%20as%3D%22targetPoint%22%2F%3E%3CArray%20as%3D%22points%22%3E%3CmxPoint%20x%3D%22-332%22%20y%3D%22280%22%2F%3E%3C%2FArray%3E%3C%2FmxGeometry%3E%3C%2FmxCell%3E%3CmxCell%20id%3D%223%22%20value%3D%22template%20author%20%E7%89%B9%E5%8C%96%22%20style%3D%22edgeLabel%3Bhtml%3D1%3Balign%3Dcenter%3BverticalAlign%3Dmiddle%3Bresizable%3D0%3Bpoints%3D%5B%5D%3B%22%20vertex%3D%221%22%20connectable%3D%220%22%20parent%3D%222%22%3E%3CmxGeometry%20x%3D%220.4298%22%20y%3D%221%22%20relative%3D%221%22%20as%3D%22geometry%22%3E%3CmxPoint%20x%3D%22-37%22%20y%3D%221%22%20as%3D%22offset%22%2F%3E%3C%2FmxGeometry%3E%3C%2FmxCell%3E%3C%2Froot%3E%3C%2FmxGraphModel%3E
%3CmxGraphModel%3E%3Croot%3E%3CmxCell%20id%3D%220%22%2F%3E%3CmxCell%20id%3D%221%22%20parent%3D%220%22%2F%3E%3CmxCell%20id%3D%222%22%20style%3D%22edgeStyle%3DorthogonalEdgeStyle%3Brounded%3D0%3BorthogonalLoop%3D1%3BjettySize%3Dauto%3Bhtml%3D1%3BentryX%3D0%3BentryY%3D0.123%3BentryDx%3D0%3BentryDy%3D0%3BentryPerimeter%3D0%3BexitX%3D1.006%3BexitY%3D0.145%3BexitDx%3D0%3BexitDy%3D0%3BexitPerimeter%3D0%3B%22%20edge%3D%221%22%20parent%3D%221%22%3E%3CmxGeometry%20relative%3D%221%22%20as%3D%22geometry%22%3E%3CmxPoint%20x%3D%22-332%22%20y%3D%22283%22%20as%3D%22sourcePoint%22%2F%3E%3CmxPoint%20x%3D%22-150%22%20y%3D%22280%22%20as%3D%22targetPoint%22%2F%3E%3CArray%20as%3D%22points%22%3E%3CmxPoint%20x%3D%22-332%22%20y%3D%22280%22%2F%3E%3C%2FArray%3E%3C%2FmxGeometry%3E%3C%2FmxCell%3E%3CmxCell%20id%3D%223%22%20value%3D%22template%20author%20%E7%89%B9%E5%8C%96%22%20style%3D%22edgeLabel%3Bhtml%3D1%3Balign%3Dcenter%3BverticalAlign%3Dmiddle%3Bresizable%3D0%3Bpoints%3D%5B%5D%3B%22%20vertex%3D%221%22%20connectable%3D%220%22%20parent%3D%222%22%3E%3CmxGeometry%20x%3D%220.4298%22%20y%3D%221%22%20relative%3D%221%22%20as%3D%22geometry%22%3E%3CmxPoint%20x%3D%22-37%22%20y%3D%221%22%20as%3D%22offset%22%2F%3E%3C%2FmxGeometry%3E%3C%2FmxCell%3E%3C%2Froot%3E%3C%2FmxGraphModel%3E
template author 精度特化
Add:TF
+ run override
Sub:TF
+ run override
 \ No newline at end of file +
executor 
ompsimd
tensorfunc<T>
tensorfunc 特化
TF
cpp-common
executor 
cuda
TFfactory
List
init.hpp
elementwise.hpp
matmul.hpp

io.hpp

reduce.hpp

changeshape.hpp
authora
init_authora.hpp
elementwise_authora.hpp
matmul_authora.hpp

io_authora.hpp

reduce_authora.hpp

changeshape_authora.hpp
authorb
matmul_authora.hpp

io_authora.hpp

reduce_authora.hpp

changeshape_authora.hpp
TF
+ name
+ args
+ returns
template author 特化
%3CmxGraphModel%3E%3Croot%3E%3CmxCell%20id%3D%220%22%2F%3E%3CmxCell%20id%3D%221%22%20parent%3D%220%22%2F%3E%3CmxCell%20id%3D%222%22%20style%3D%22edgeStyle%3DorthogonalEdgeStyle%3Brounded%3D0%3BorthogonalLoop%3D1%3BjettySize%3Dauto%3Bhtml%3D1%3BentryX%3D0%3BentryY%3D0.123%3BentryDx%3D0%3BentryDy%3D0%3BentryPerimeter%3D0%3BexitX%3D1.006%3BexitY%3D0.145%3BexitDx%3D0%3BexitDy%3D0%3BexitPerimeter%3D0%3B%22%20edge%3D%221%22%20parent%3D%221%22%3E%3CmxGeometry%20relative%3D%221%22%20as%3D%22geometry%22%3E%3CmxPoint%20x%3D%22-332%22%20y%3D%22283%22%20as%3D%22sourcePoint%22%2F%3E%3CmxPoint%20x%3D%22-150%22%20y%3D%22280%22%20as%3D%22targetPoint%22%2F%3E%3CArray%20as%3D%22points%22%3E%3CmxPoint%20x%3D%22-332%22%20y%3D%22280%22%2F%3E%3C%2FArray%3E%3C%2FmxGeometry%3E%3C%2FmxCell%3E%3CmxCell%20id%3D%223%22%20value%3D%22template%20author%20%E7%89%B9%E5%8C%96%22%20style%3D%22edgeLabel%3Bhtml%3D1%3Balign%3Dcenter%3BverticalAlign%3Dmiddle%3Bresizable%3D0%3Bpoints%3D%5B%5D%3B%22%20vertex%3D%221%22%20connectable%3D%220%22%20parent%3D%222%22%3E%3CmxGeometry%20x%3D%220.4298%22%20y%3D%221%22%20relative%3D%221%22%20as%3D%22geometry%22%3E%3CmxPoint%20x%3D%22-37%22%20y%3D%221%22%20as%3D%22offset%22%2F%3E%3C%2FmxGeometry%3E%3C%2FmxCell%3E%3C%2Froot%3E%3C%2FmxGraphModel%3E
%3CmxGraphModel%3E%3Croot%3E%3CmxCell%20id%3D%220%22%2F%3E%3CmxCell%20id%3D%221%22%20parent%3D%220%22%2F%3E%3CmxCell%20id%3D%222%22%20style%3D%22edgeStyle%3DorthogonalEdgeStyle%3Brounded%3D0%3BorthogonalLoop%3D1%3BjettySize%3Dauto%3Bhtml%3D1%3BentryX%3D0%3BentryY%3D0.123%3BentryDx%3D0%3BentryDy%3D0%3BentryPerimeter%3D0%3BexitX%3D1.006%3BexitY%3D0.145%3BexitDx%3D0%3BexitDy%3D0%3BexitPerimeter%3D0%3B%22%20edge%3D%221%22%20parent%3D%221%22%3E%3CmxGeometry%20relative%3D%221%22%20as%3D%22geometry%22%3E%3CmxPoint%20x%3D%22-332%22%20y%3D%22283%22%20as%3D%22sourcePoint%22%2F%3E%3CmxPoint%20x%3D%22-150%22%20y%3D%22280%22%20as%3D%22targetPoint%22%2F%3E%3CArray%20as%3D%22points%22%3E%3CmxPoint%20x%3D%22-332%22%20y%3D%22280%22%2F%3E%3C%2FArray%3E%3C%2FmxGeometry%3E%3C%2FmxCell%3E%3CmxCell%20id%3D%223%22%20value%3D%22template%20author%20%E7%89%B9%E5%8C%96%22%20style%3D%22edgeLabel%3Bhtml%3D1%3Balign%3Dcenter%3BverticalAlign%3Dmiddle%3Bresizable%3D0%3Bpoints%3D%5B%5D%3B%22%20vertex%3D%221%22%20connectable%3D%220%22%20parent%3D%222%22%3E%3CmxGeometry%20x%3D%220.4298%22%20y%3D%221%22%20relative%3D%221%22%20as%3D%22geometry%22%3E%3CmxPoint%20x%3D%22-37%22%20y%3D%221%22%20as%3D%22offset%22%2F%3E%3C%2FmxGeometry%3E%3C%2FmxCell%3E%3C%2Froot%3E%3C%2FmxGraphModel%3E
template author 精度特化
Add:TF
+ run override
Sub:TF
+ run override
 \ No newline at end of file diff --git a/docs/executor/excuter.md b/docs/executor/executor.md similarity index 81% rename from docs/executor/excuter.md rename to docs/executor/executor.md index 0e604e04..0a5b4633 100644 --- a/docs/executor/excuter.md +++ b/docs/executor/executor.md @@ -1,4 +1,4 @@ -## 如何给excuter添加一个新算子 +## 如何给executor添加一个新算子 ### 层次结构图 @@ -14,7 +14,7 @@ #### Op -Op是excuter的算子，是excuter的执行单元 +Op是executor的算子，是executor的执行单元 在程序中，Op是基类，不同的Op有不同的实现，比如Add, Mul, MatMul等。 每个Op都需要override forward和backward函数 @@ -31,7 +31,7 @@ Matmul会选择选择一个默认的实现 git clone https://github.com/deepx-org/deepx.git #### 1.cpu执行器 -cd deepx/excuter/op-mem-ompsimd +cd deepx/executor/op-mem-ompsimd 需要提前安装好依赖 + highway需要源码安装 @@ -43,7 +43,7 @@ make build && cd build && cmake .. && make #### 2.cuda执行器 -cd deepx/excuter/op-mem-cuda +cd deepx/executor/op-mem-cuda 需要提前安装好依赖 + cuda @@ -60,7 +60,7 @@ todo #### 4.front对接测试 -1.先启动excuter可执行文件， 位于excuter/op-mem-{cuda/ompsimd}/build，可执行文件名同excuter名 +1.先启动executor可执行文件， 位于executor/op-mem-{cuda/ompsimd}/build，可执行文件名同executor名 2.然后测试front中py的对应算子脚本（front/py/examples 目录） 可以按照顺序，以此测试 diff --git a/docs/executor/mix_precision.md b/docs/executor/mix_precision.md index 1ca04687..ef2941cb 100644 --- a/docs/executor/mix_precision.md +++ b/docs/executor/mix_precision.md @@ -8,7 +8,7 @@ mix precision 是一种混合精度训练方法，它使用 16 位浮点数和 8 在深度学习中，模型通常使用 32 位浮点数进行训练，这样可以确保模型的精度。但是，32 位浮点数占用的显存较大，计算时间较长。因此，为了减少显存占用和计算时间，可以使用 mix precision 训练方法。 -## 3. 关于excuter的mix precision的实现 +## 3. 关于executor的mix precision的实现 如： diff --git a/docs/executor/op-mem-cuda/list.md b/docs/executor/op-mem-cuda/list.md index 69966acf..80f19084 100644 --- a/docs/executor/op-mem-cuda/list.md +++ b/docs/executor/op-mem-cuda/list.md @@ -1,6 +1,6 @@ ## op-mem-cuda 支持算子列表 -本页面由 `excuter/op-mem-cuda 生成，请勿手动修改 +本页面由 `executor/op-mem-cuda 生成，请勿手动修改 ### matmul diff --git a/docs/executor/op-mem-ompsimd/contribute.md b/docs/executor/op-mem-ompsimd/contribute.md index b8e095c4..7e876738 100644 --- a/docs/executor/op-mem-ompsimd/contribute.md +++ b/docs/executor/op-mem-ompsimd/contribute.md @@ -1,4 +1,4 @@ -## excuter +## executor ### op-mem-ompsimd diff --git a/docs/executor/op-mem-ompsimd/list.md b/docs/executor/op-mem-ompsimd/list.md index 3ec2089e..6313a1da 100644 --- a/docs/executor/op-mem-ompsimd/list.md +++ b/docs/executor/op-mem-ompsimd/list.md @@ -1,6 +1,6 @@ ## op-mem-ompsimd 支持算子列表 -本页面由 `excuter/op-mem-ompsimd 生成，请勿手动修改 +本页面由 `executor/op-mem-ompsimd 生成，请勿手动修改 ### matmul diff --git a/docs/executor/op-mem-ompsimd/range.md b/docs/executor/op-mem-ompsimd/range.md index b0801634..ea1dd0ea 100644 --- a/docs/executor/op-mem-ompsimd/range.md +++ b/docs/executor/op-mem-ompsimd/range.md @@ -1,4 +1,4 @@ -## excuter +## executor ### op-mem-ompsimd @@ -8,9 +8,9 @@ range函数是shape类中的一个函数，用于根据shape对tensor进行omp 定义和实现分别在： -excuter/common/src/deepx/shape.hpp +executor/common/src/deepx/shape.hpp -excuter/common/src/deepx/shape_range.cpp +executor/common/src/deepx/shape_range.cpp | func | omp并行 | omp线程local局部对象 | 调用场景 | | ---- | ---- | ------ | ---------- | diff --git a/docs/front/deepx.op.drawio.svg b/docs/front/deepx.op.drawio.svg index 8c207b90..038c875f 100644 --- a/docs/front/deepx.op.drawio.svg +++ b/docs/front/deepx.op.drawio.svg @@ -1,4 +1,4 @@ -
excuter 
cuda
excuter 
cpu
excuter 
cuda
class Op{
name string;
vector<string> args;
vector<string> returns;
bool grad;
vector<string> args_grad;
vector<string> returns_grad;

virusal void forward();
virsual void backward();
}

openblas的matmul实现
template <T >
class Matmul_cblas:Op
cblas实现

tensorfunc/elementwise.hpp

template<T>
void matmul_cblas(const Tensor<T> &a, const Tensor<T> &b, Tensor<T> &c)

cblas对double的特化

template <>
void matmul<double>(const Tensor<double> &a, const Tensor<double> &b, Tensor<double> &c)
某auther a的实现
template <T >
class Matmul_authora:Op
默认的matmul
会选其中一个作为默认
template <T >
class Matmul:Op
某auther a的实现

tensorfunc/elementwise.hpp

template<T>
void matmul_authora(const Tensor<T> &a, const Tensor<T> &b, Tensor<T> &c)

authora对double的特化

template <>
  void matmul_authora<double>(const Tensor<double> &a, const Tensor<double> &b, Tensor<double> &c)
tensorfunc<T>
tensorfunc 特化
Op
 \ No newline at end of file +
executor 
cuda
executor 
cpu
executor 
cuda
class Op{
name string;
vector<string> args;
vector<string> returns;
bool grad;
vector<string> args_grad;
vector<string> returns_grad;

virusal void forward();
virsual void backward();
}

openblas的matmul实现
template <T >
class Matmul_cblas:Op
cblas实现

tensorfunc/elementwise.hpp

template<T>
void matmul_cblas(const Tensor<T> &a, const Tensor<T> &b, Tensor<T> &c)

cblas对double的特化

template <>
void matmul<double>(const Tensor<double> &a, const Tensor<double> &b, Tensor<double> &c)
某auther a的实现
template <T >
class Matmul_authora:Op
默认的matmul
会选其中一个作为默认
template <T >
class Matmul:Op
某auther a的实现

tensorfunc/elementwise.hpp

template<T>
void matmul_authora(const Tensor<T> &a, const Tensor<T> &b, Tensor<T> &c)

authora对double的特化

template <>
  void matmul_authora<double>(const Tensor<double> &a, const Tensor<double> &b, Tensor<double> &c)
tensorfunc<T>
tensorfunc 特化
Op
 \ No newline at end of file diff --git a/docs/index.rst b/docs/index.rst index 0162c9f7..1566faf5 100644 --- a/docs/index.rst +++ b/docs/index.rst @@ -18,7 +18,7 @@ DeepX 原生分布式并行的深度学习训练推理一体框架 :caption: doc 文档 front/py/deepx/about - excuter/op-mem-ompsimd/list + executor/op-mem-ompsimd/list deepxIR/ir .. toctree:: @@ -27,9 +27,9 @@ DeepX 原生分布式并行的深度学习训练推理一体框架 front/py/contribute scheduler/scheduler - excuter/excuter - excuter/op-mem-ompsimd/contribute - excuter/op-mem-ompsimd/range + executor/executor + executor/op-mem-ompsimd/contribute + executor/op-mem-ompsimd/range 索引和搜索 ========== diff --git a/docs/language.md b/docs/language.md index 0a78594d..0e8a5f72 100644 --- a/docs/language.md +++ b/docs/language.md @@ -1,4 +1,4 @@ -## c++:计算执行器(excuter) +## c++:计算执行器(executor) 负责实现tensor的具体计算过程,对接硬件如GPU、CPU的simd指令 @@ -33,9 +33,9 @@ deepxctl:提供对deepx体系的所有工具、库、模型、镜像的统一纳 ## deepxIR 虽然deepxIR不是独立的编程语言，但是deepx体系的程序格式标准 -excuter所执行的内容，就是deepxir的序列或deepxir计算图 +executor所执行的内容，就是deepxir的序列或deepxir计算图 -https://github.com/array2d/deepx/blob/main/doc/excuter/op-mem-cuda/list.md +https://github.com/array2d/deepx/blob/main/doc/executor/op-mem-cuda/list.md deepxir分为4类 diff --git a/docs/scheduler/scheduler.md b/docs/scheduler/scheduler.md index c1d9cf0a..5f7d258c 100644 --- a/docs/scheduler/scheduler.md +++ b/docs/scheduler/scheduler.md @@ -1,12 +1,12 @@ ### scheduler -DeepX框架的scheduler，是front和excuter之间的桥梁。 +DeepX框架的scheduler，是front和executor之间的桥梁。 -front只负责搭建抽象计算图，excuter负责执行算子，而scheduler负责将抽象计算图转换为执行计算图，并发送给excuter。 +front只负责搭建抽象计算图，executor负责执行算子，而scheduler负责将抽象计算图转换为执行计算图，并发送给executor。 #### 算子注册器 -算子注册器，接收excuter的算子及精度列表。 +算子注册器，接收executor的算子及精度列表。 #### 调度器 diff --git a/executor/cpp-common/src/deepx/tf/tffactory.cpp b/executor/cpp-common/src/deepx/tf/tffactory.cpp index c57adf51..7ddbf555 100644 --- a/executor/cpp-common/src/deepx/tf/tffactory.cpp +++ b/executor/cpp-common/src/deepx/tf/tffactory.cpp @@ -82,11 +82,11 @@ namespace deepx::tf cloned->metadata=other.metadata; return cloned; } - string TfFactory::print_markdown(string excuter_name) const + string TfFactory::print_markdown(string executor_name) const { std::stringstream ss; - ss << "## " << excuter_name << " 支持算子列表 \n\n"; - ss << "本页面由 `excuter/" << excuter_name << " 生成，请勿手动修改 \n\n"; + ss << "## " << executor_name << " 支持算子列表 \n\n"; + ss << "本页面由 `executor/" << executor_name << " 生成，请勿手动修改 \n\n"; // 首先按tftype分组 unordered_map>> tf_by_type; diff --git a/executor/cpp-common/src/deepx/tf/tffactory.hpp b/executor/cpp-common/src/deepx/tf/tffactory.hpp index ebe836c0..6f8e97dc 100644 --- a/executor/cpp-common/src/deepx/tf/tffactory.hpp +++ b/executor/cpp-common/src/deepx/tf/tffactory.hpp @@ -96,7 +96,7 @@ namespace deepx::tf } shared_ptr get_tf(const TF &other) const; // 输出为markdown表格格式 - string print_markdown(string excuter_name) const; + string print_markdown(string executor_name) const; }; } diff --git a/executor/op-mem-cuda/CMakeLists.txt b/executor/op-mem-cuda/CMakeLists.txt index 98471042..f373c08e 100644 --- a/executor/op-mem-cuda/CMakeLists.txt +++ b/executor/op-mem-cuda/CMakeLists.txt @@ -1,5 +1,5 @@ cmake_minimum_required(VERSION 3.15...3.29) -project(deepx-excuter-cuda LANGUAGES CXX CUDA) +project(deepx-executor-cuda LANGUAGES CXX CUDA) # 设置 C++ 标准 set(CMAKE_CXX_STANDARD 17) diff --git a/executor/op-mem-cuda/Dockerfile b/executor/op-mem-cuda/Dockerfile index 682b6fd6..e78e4055 100644 --- a/executor/op-mem-cuda/Dockerfile +++ b/executor/op-mem-cuda/Dockerfile @@ -20,4 +20,4 @@ RUN rm -rf build && mkdir build && cd build && \ cmake .. && \ make -j$(nproc) -CMD ["./build/bin/deepx-excuter-cuda"] \ No newline at end of file +CMD ["./build/bin/deepx-executor-cuda"] \ No newline at end of file diff --git a/executor/op-mem-cuda/README.md b/executor/op-mem-cuda/README.md index c0014d63..08efee23 100644 --- a/executor/op-mem-cuda/README.md +++ b/executor/op-mem-cuda/README.md @@ -8,7 +8,7 @@ sudo bash build.sh ### docker 构建命令，及所在位置 ``` -lipeng@lp:~/code/git.array2d.com/ai/deepx/excuter/op-mem-cuda$ sudo bash dockerbuild.sh +lipeng@lp:~/code/git.array2d.com/ai/deepx/executor/op-mem-cuda$ sudo bash dockerbuild.sh ``` \ No newline at end of file diff --git a/executor/op-mem-cuda/src/client/main.cpp b/executor/op-mem-cuda/src/client/main.cpp index b00a284a..a198d831 100644 --- a/executor/op-mem-cuda/src/client/main.cpp +++ b/executor/op-mem-cuda/src/client/main.cpp @@ -29,7 +29,7 @@ int main() register_all(tf_factory); // 将op table输出到markdown文件 - string docdir = "../../../doc/excuter/op-mem-cuda/"; + string docdir = "../../../doc/executor/op-mem-cuda/"; std::ofstream md_file(docdir + "list.md"); if (md_file.is_open()) { diff --git a/executor/op-mem-ompsimd/.cursorrules b/executor/op-mem-ompsimd/.cursorrules index fa3e3454..9c2ede76 100644 --- a/executor/op-mem-ompsimd/.cursorrules +++ b/executor/op-mem-ompsimd/.cursorrules @@ -1 +1 @@ -excuter只实现基础的op，不要实现可以由基础op组合的op，如relu可以由max组合。 \ No newline at end of file +executor只实现基础的op，不要实现可以由基础op组合的op，如relu可以由max组合。 \ No newline at end of file diff --git a/executor/op-mem-ompsimd/CMakeLists.txt b/executor/op-mem-ompsimd/CMakeLists.txt index daab9fd0..abc1a868 100644 --- a/executor/op-mem-ompsimd/CMakeLists.txt +++ b/executor/op-mem-ompsimd/CMakeLists.txt @@ -1,5 +1,5 @@ cmake_minimum_required(VERSION 3.15...3.29) -project(deepx-excuter-ompsimd LANGUAGES CXX ) +project(deepx-executor-ompsimd LANGUAGES CXX ) # 设置 C++ 标准 set(CMAKE_CXX_STANDARD 17) diff --git a/executor/op-mem-ompsimd/Dockerfile b/executor/op-mem-ompsimd/Dockerfile index acd79b6f..1b4fd775 100644 --- a/executor/op-mem-ompsimd/Dockerfile +++ b/executor/op-mem-ompsimd/Dockerfile @@ -31,4 +31,4 @@ RUN rm -rf build && mkdir build && cd build && \ cmake ..&& \ make -j$(nproc) -CMD ["./build/bin/deepx-excuter-ompsimd"] \ No newline at end of file +CMD ["./build/bin/deepx-executor-ompsimd"] \ No newline at end of file diff --git a/executor/op-mem-ompsimd/src/client/main.cpp b/executor/op-mem-ompsimd/src/client/main.cpp index 15004184..b0c79fe4 100644 --- a/executor/op-mem-ompsimd/src/client/main.cpp +++ b/executor/op-mem-ompsimd/src/client/main.cpp @@ -30,7 +30,7 @@ int main() register_all(tf_factory); // 将op table输出到markdown文件 - string docdir = "../../../doc/excuter/op-mem-ompsimd/"; + string docdir = "../../../doc/executor/op-mem-ompsimd/"; std::ofstream md_file(docdir + "list.md"); if (md_file.is_open()) { diff --git a/front/go/README.md b/front/go/README.md index bd2e1da0..72142e9d 100644 --- a/front/go/README.md +++ b/front/go/README.md @@ -8,7 +8,7 @@ deepx-py是控制侧，负责生成计算图，通过通信调用其他进程 -在excuter中，实现了具体的存算引擎 +在executor中，实现了具体的存算引擎 diff --git a/front/py/.cursorrules b/front/py/.cursorrules index 91c6c0ef..42747b69 100644 --- a/front/py/.cursorrules +++ b/front/py/.cursorrules @@ -9,6 +9,6 @@ Always respond in 中文 是deepx的python部分，项目路径为/home/lipeng/code/ai/deepx/front/py deepx的对外暴露的设计，尽可能接近pytorch的API deepx_py是前端，负责搭建深度学习模型,但并不参与实际数据存储和计算 -deepx_py虽然有类似pytorch的API,但只是调用excuter内的对应实现 -deepx_py依赖网络api调用后端excuter的计算 -deepx_py的每个OP，都会在excuter内找到对应的实现 \ No newline at end of file +deepx_py虽然有类似pytorch的API,但只是调用executor内的对应实现 +deepx_py依赖网络api调用后端executor的计算 +deepx_py的每个OP，都会在executor内找到对应的实现 \ No newline at end of file diff --git a/front/py/deepx/README.md b/front/py/deepx/README.md index 5e131e97..979d09ac 100644 --- a/front/py/deepx/README.md +++ b/front/py/deepx/README.md @@ -8,7 +8,7 @@ deepx-py是为了构建抽象计算图的py接口库 deepx-py是控制侧，负责生成计算图，通过通信调用其他进程 -在excuter中，实现了具体的存算引擎 +在executor中，实现了具体的存算引擎 diff --git a/rename.py b/rename.py new file mode 100644 index 00000000..e8ee68f8 --- /dev/null +++ b/rename.py @@ -0,0 +1,18 @@ +#!/usr/bin/env python3 +import subprocess, pathlib, sys + +paths = subprocess.check_output( + ["git", "ls-files", "-z"], text=True +).split("\0") + +print(paths) + +# 倒序，先文件再目录 +for p in sorted(filter(None, paths), reverse=True): + if "excuter" in p: + print(f"find: {p}\n") + new = p.replace("excuter", "executor") + + subprocess.run(["git", "mv", p, new], check=True) + +print("Done. Now run your tests.") From b272c82e8204e6d51b9986ed01351930b362910e Mon Sep 17 00:00:00 2001 From: LJXjean <1372682462@qq.com> Date: Sat, 5 Jul 2025 16:03:10 +0200 Subject: [PATCH 3/6] doc to docs --- README.md | 4 +- all_excuter.txt | 297 ------------------ docs/README.md | 2 +- docs/language.md | 2 +- .../{doc => docs}/00_quickstart.md | 0 .../op-mem-cuda/{doc => docs}/01_layout.md | 0 .../{doc => docs}/02_layout_algebra.md | 0 .../op-mem-cuda/{doc => docs}/03_tensor.md | 0 .../{doc => docs}/04_algorithms.md | 0 .../op-mem-cuda/{doc => docs}/0t_mma_atom.md | 0 .../{doc => docs}/0x_gemm_tutorial.md | 0 .../{doc => docs}/0y_predication.md | 0 .../{doc => docs}/0z_tma_tensors.md | 0 executor/op-mem-cuda/src/client/main.cpp | 4 +- executor/op-mem-ompsimd/src/client/main.cpp | 4 +- 15 files changed, 8 insertions(+), 305 deletions(-) delete mode 100644 all_excuter.txt rename executor/op-mem-cuda/{doc => docs}/00_quickstart.md (100%) rename executor/op-mem-cuda/{doc => docs}/01_layout.md (100%) rename executor/op-mem-cuda/{doc => docs}/02_layout_algebra.md (100%) rename executor/op-mem-cuda/{doc => docs}/03_tensor.md (100%) rename executor/op-mem-cuda/{doc => docs}/04_algorithms.md (100%) rename executor/op-mem-cuda/{doc => docs}/0t_mma_atom.md (100%) rename executor/op-mem-cuda/{doc => docs}/0x_gemm_tutorial.md (100%) rename executor/op-mem-cuda/{doc => docs}/0y_predication.md (100%) rename executor/op-mem-cuda/{doc => docs}/0z_tma_tensors.md (100%) diff --git a/README.md b/README.md index a24a727f..50a06985 100644 --- a/README.md +++ b/README.md @@ -50,10 +50,10 @@ Op需要实现run方法 #### 默认执行器 -+ cpu执行器,已实现ompsimd。其支持的算子列表[ompsimd](doc/executor/op-mem-ompsimd/list.md) ++ cpu执行器,已实现ompsimd。其支持的算子列表[ompsimd](docs/executor/op-mem-ompsimd/list.md) #### GPU执行器 -+ cuda执行器，其支持的算子列表[cuda](doc/executor/op-mem-cuda/list.md) ++ cuda执行器，其支持的算子列表[cuda](docs/executor/op-mem-cuda/list.md) 欢迎大家提交cuda代码 diff --git a/all_excuter.txt b/all_excuter.txt deleted file mode 100644 index 341f230f..00000000 --- a/all_excuter.txt +++ /dev/null @@ -1,297 +0,0 @@ -./README.md:18:+ IR通信调度。不同于pytorch或其他py+bind c++这种单一进程的栈上函数调度执行的方式。deepx各个程序（如front的python sdk，back的计算图编译器优化器、executor如ompsimd）之间，通过IR实现网络通信调度，需要各自启动对应进程。 -./README.md:33:+ front生成基础IR，编译器负责进行fusion成executor注册的高级算子。 -./README.md:47:关于executor，只要能按deepxIR序列执行，并返回结果，就可以接入deepx分布式调度框架，因此，从硬件、指令、加速库、高级框架包括训练、推理引擎，都可以稍作修改，就接入deepx体系。 -./README.md:53:+ cpu执行器,已实现ompsimd。其支持的算子列表[ompsimd](doc/executor/op-mem-ompsimd/list.md) -./README.md:56:+ cuda执行器，其支持的算子列表[cuda](doc/executor/op-mem-cuda/list.md) -./docs/language.md:1:## c++:计算执行器(executor) -./docs/language.md:36:executor所执行的内容，就是deepxir的序列或deepxir计算图 -./docs/language.md:38:https://github.com/array2d/deepx/blob/main/doc/executor/op-mem-cuda/list.md -./docs/scheduler/scheduler.md:3:DeepX框架的scheduler，是front和executor之间的桥梁。 -./docs/scheduler/scheduler.md:5:front只负责搭建抽象计算图，executor负责执行算子，而scheduler负责将抽象计算图转换为执行计算图，并发送给executor。 -./docs/scheduler/scheduler.md:9:算子注册器，接收executor的算子及精度列表。 -./docs/executor/mix_precision.md:11:## 3. 关于executor的mix precision的实现 -./docs/executor/op-mem-cuda/list.md:3:本页面由 `executor/op-mem-cuda 生成，请勿手动修改 -./docs/executor/deepx.op.drawio.svg:4:
executor 
ompsimd
tensorfunc<T>
tensorfunc 特化
TF
cpp-common
executor 
cuda
 -./docs/executor/deepx.op.drawio:28: -./docs/executor/deepx.op.drawio:135: -./docs/executor/executor.md:1:## 如何给executor添加一个新算子 -./docs/executor/executor.md:17:Op是executor的算子，是executor的执行单元 -./docs/executor/executor.md:34:cd deepx/executor/op-mem-ompsimd -./docs/executor/executor.md:46:cd deepx/executor/op-mem-cuda -./docs/executor/executor.md:63:1.先启动executor可执行文件， 位于executor/op-mem-{cuda/ompsimd}/build，可执行文件名同executor名 -./docs/design.md:18:executor执行deepxIR的规则 -./docs/design.md:20:+ executor执行deepxIR时，不得修改args中的tensor -./docs/front/deepx.op.drawio.svg:4:
executor 
cuda
executor 
cpu
executor 
cuda
 Date: Sat, 5 Jul 2025 16:03:52 +0200 Subject: [PATCH 4/6] remove util files --- all_doc | 7 ------- rename.py | 18 ------------------ 2 files changed, 25 deletions(-) delete mode 100644 all_doc delete mode 100644 rename.py diff --git a/all_doc b/all_doc deleted file mode 100644 index 83dc5613..00000000 --- a/all_doc +++ /dev/null @@ -1,7 +0,0 @@ -README.md:53:+ cpu执行器,已实现ompsimd。其支持的算子列表[ompsimd](doc/executor/op-mem-ompsimd/list.md) -README.md:56:+ cuda执行器，其支持的算子列表[cuda](doc/executor/op-mem-cuda/list.md) -docs/README.md:56:![架构图](./doc/deepx.jpg) -docs/front/deepxpy.drawio.svg:4:
autograd
autograd
functional

elementwise:add,sub..
matmul:
reduce:sum,min,max

functional...
tensor

shape
dtype
tensor...
OpNode
OpNode
DataNode
type:var,vector,tensor
data:
DataNode...
Node
Node
Graph
- nodes
Graph...
deepx-front-py
deepx-front-py
ControlflowNode
ControlflowNode
nn
nn
deepxIR
deepxIRText is not SVG - cannot display -docs/language.md:38:https://github.com/array2d/deepx/blob/main/doc/executor/op-mem-cuda/list.md -executor/op-mem-cuda/src/client/main.cpp:32: string docdir = "../../../doc/executor/op-mem-cuda/"; -executor/op-mem-ompsimd/src/client/main.cpp:33: string docdir = "../../../doc/executor/op-mem-ompsimd/"; diff --git a/rename.py b/rename.py deleted file mode 100644 index e8ee68f8..00000000 --- a/rename.py +++ /dev/null @@ -1,18 +0,0 @@ -#!/usr/bin/env python3 -import subprocess, pathlib, sys - -paths = subprocess.check_output( - ["git", "ls-files", "-z"], text=True -).split("\0") - -print(paths) - -# 倒序，先文件再目录 -for p in sorted(filter(None, paths), reverse=True): - if "excuter" in p: - print(f"find: {p}\n") - new = p.replace("excuter", "executor") - - subprocess.run(["git", "mv", p, new], check=True) - -print("Done. Now run your tests.") From 77663b30a925b0a6739b4a966920981db12746f6 Mon Sep 17 00:00:00 2001 From: LJXjean <1372682462@qq.com> Date: Thu, 10 Jul 2025 13:33:10 +0200 Subject: [PATCH 5/6] docker modified for testing --- cutlass | 1 + executor/op-mem-cuda/Dockerfile | 8 ++++---- 2 files changed, 5 insertions(+), 4 deletions(-) create mode 160000 cutlass diff --git a/cutlass b/cutlass new file mode 160000 index 00000000..bbe579a9 --- /dev/null +++ b/cutlass @@ -0,0 +1 @@ +Subproject commit bbe579a9e3beb6ea6626d9227ec32d0dae119a49 diff --git a/executor/op-mem-cuda/Dockerfile b/executor/op-mem-cuda/Dockerfile index e78e4055..60fda776 100644 --- a/executor/op-mem-cuda/Dockerfile +++ b/executor/op-mem-cuda/Dockerfile @@ -10,14 +10,14 @@ RUN apt-get update && apt-get install -y \ && rm -rf /var/lib/apt/lists/* -WORKDIR /home/ +WORKDIR /home # CUDA 已经预装，无需额外配置 -ADD cpp-common cpp-common -ADD op-mem-cuda op-mem-cuda +ADD executor/cpp-common cpp-common +ADD executor/op-mem-cuda op-mem-cuda WORKDIR /home/op-mem-cuda RUN rm -rf build && mkdir build && cd build && \ cmake .. && \ make -j$(nproc) -CMD ["./build/bin/deepx-executor-cuda"] \ No newline at end of file +ENTRYPOINT ["./build/deepx-executor-cuda"] \ No newline at end of file From 1f3eafbb7a899a1646a0c948935aa434d1cbb4d5 Mon Sep 17 00:00:00 2001 From: LJXjean <1372682462@qq.com> Date: Thu, 10 Jul 2025 15:37:43 +0200 Subject: [PATCH 6/6] fix .github/workflows --- .github/workflows/executor-cppcommon.yml | 2 +- .github/workflows/executor-cuda-linux.yml | 2 +- .github/workflows/executor-ompsimd-linux.yml | 3 +-- .github/workflows/tool-deepxctl.yml | 2 +- 4 files changed, 4 insertions(+), 5 deletions(-) diff --git a/.github/workflows/executor-cppcommon.yml b/.github/workflows/executor-cppcommon.yml index efa39092..3b0e5794 100644 --- a/.github/workflows/executor-cppcommon.yml +++ b/.github/workflows/executor-cppcommon.yml @@ -1,5 +1,5 @@ name: Excuter/cppcommon Build -on: [push, pull_request] +on: push: paths: - 'executor/cpp-common/**' diff --git a/.github/workflows/executor-cuda-linux.yml b/.github/workflows/executor-cuda-linux.yml index 4bf3359c..a075ffa5 100644 --- a/.github/workflows/executor-cuda-linux.yml +++ b/.github/workflows/executor-cuda-linux.yml @@ -1,5 +1,5 @@ name: Excuter/cuda-linux Build -on: [push, pull_request] +on: push: paths: - 'executor/op-mem-cuda/**' diff --git a/.github/workflows/executor-ompsimd-linux.yml b/.github/workflows/executor-ompsimd-linux.yml index 4b4d08b7..33f3cb11 100644 --- a/.github/workflows/executor-ompsimd-linux.yml +++ b/.github/workflows/executor-ompsimd-linux.yml @@ -1,6 +1,5 @@ name: Excuter/ompsimd-linux Build -on: [push, pull_request] -on: [push, pull_request] +on: push: paths: - 'executor/op-mem-ompsimd/**' diff --git a/.github/workflows/tool-deepxctl.yml b/.github/workflows/tool-deepxctl.yml index 39d3cb9e..eee66b31 100644 --- a/.github/workflows/tool-deepxctl.yml +++ b/.github/workflows/tool-deepxctl.yml @@ -1,5 +1,5 @@ name: Tool/deepxctl Build -on: [push, pull_request] +on: push: paths: - 'tool/deepxctl/**'