After lowering is done, ``build()`` function generates target machine code from the lowered function. This code can contain SSE or AVX instructions if you target x86, or PTX instructions for CUDA target. In addition to target specific machine code, TVM also generates host side code that is responsible for memory management, kernel launch etc.
-Code generation is done by ``build_module()`` function, defined in ``python/tvm/codege.py``. On the C++ side, code generation is implemented in ``src/codegen`` subdirectory. ``build_module()`` Python function will reach ``Build()`` function below in ``src/codegen/codegen.cc``:
+Code generation is done by ``build_module()`` function, defined in ``python/tvm/codegen.py``. On the C++ side, code generation is implemented in ``src/codegen`` subdirectory. ``build_module()`` Python function will reach ``Build()`` function below in ``src/codegen/codegen.cc``:
::
# KIND, either express or implied. See the License for the
# specific language governing permissions and limitations
# under the License.
-"""Backend codege modules for relay."""
+"""Backend codegen modules for relay."""
from . import compile_engine