This is an automated email from the ASF dual-hosted git repository.
junrushao pushed a change to branch v0.8
in repository https://gitbox.apache.org/repos/asf/tvm.git.
from 86781e9 [ETHOSN] Streamline Ethos(TM)-N cross-compile rpc usage
(#9477)
add 0812c07 Change Call with TIRCallAttrs to call_lowered op (#9312)
add dc56eea [Support] Add libinfo into the runtime build (#9310)
add f1c2c5f Fixed some warnings about lambda's closures that are bigger
than necessary (#9481)
add b26ddfe [HOTFIX][TARGET] Change LOG in compilation config to DLOG
(#9486)
add 1e09bb2 Add default for split op (#9489)
add f7b3909 [COMMUNITY] Junru's and Wuwei's PGP key for ASF release
(#9488)
add e81d391 Arm(R) Ethos(TM)-U NPU BinaryElementwise operators support
(#9442)
add e5427b2 [Relay] Use target_host determined at Relay level instead of
recalculating it (#9499)
add 7f3d363 Add LLVM-13 installation to Docker setup (#9498)
add 4bebfd8 Fix repository URL in ubuntu_install_rocm.sh (#9425)
add 6159b8e [Topi][Op][PyTorch][Vitas] Fix inconsistent kernel layout
conventions for conv2d_transpose (#9336)
add 137def8 [TARGET] Move target_host usage to new target style. (#9497)
add 2dad4d1 [BugFix] Fix divide by zero error in TIR pass
lower_warp_memory (#9485)
add 3ad7c4a [Conv2DTransposed] Fix wrong shape check and add new TOPI
module to support groups (#9465)
No new revisions were added by this update.
Summary of changes:
CMakeLists.txt | 4 +-
KEYS | 116 ++++++
apps/android_camera/models/prepare_model.py | 2 +-
apps/android_rpc/tests/android_rpc_test.py | 4 +-
apps/benchmark/arm_cpu_imagenet_bench.py | 2 +-
apps/benchmark/mobile_gpu_imagenet_bench.py | 2 +-
apps/hexagon_launcher/README.md | 18 +-
apps/ios_rpc/tests/ios_rpc_mobilenet.py | 4 +-
apps/ios_rpc/tests/ios_rpc_test.py | 2 +-
apps/topi_recipe/gemm/android_gemm_square.py | 2 +-
docker/install/ubuntu1804_install_llvm.sh | 14 +-
docker/install/ubuntu_install_rocm.sh | 6 +-
docs/how_to/deploy/bnns.rst | 2 +-
docs/how_to/deploy/hls.rst | 5 +-
golang/sample/deploy.py | 3 +-
include/tvm/relay/attrs/annotation.h | 11 -
include/tvm/relay/attrs/{debug.h => call.h} | 22 +-
jvm/README.md | 2 +-
jvm/core/src/test/scripts/test_add_cpu.py | 2 +-
jvm/core/src/test/scripts/test_add_gpu.py | 2 +-
python/tvm/__init__.py | 5 +-
python/tvm/auto_scheduler/relay_integration.py | 6 +
python/tvm/autotvm/task/relay_integration.py | 6 +
python/tvm/driver/build_module.py | 9 +-
python/tvm/micro/__init__.py | 7 +-
.../tvm/relay/backend/contrib/ethosu/legalize.py | 227 +++++++++++
.../relay/backend/contrib/ethosu/op/__init__.py | 1 +
.../contrib/ethosu/op/binary_elementwise.py | 215 ++++++++++
.../relay/backend/contrib/ethosu/op/convolution.py | 4 +-
.../relay/backend/contrib/ethosu/op/depthwise.py | 7 +-
.../tvm/relay/backend/contrib/ethosu/op/pooling.py | 6 +-
.../relay/backend/contrib/ethosu/te/__init__.py | 1 +
.../contrib/ethosu/te/binary_elementwise.py | 184 +++++++++
.../tir/{pooling.py => binary_elementwise.py} | 74 ++--
.../tvm/relay/backend/contrib/ethosu/tir/passes.py | 2 +
.../tvm/relay/backend/contrib/ethosu/tir/spec.py | 21 +
.../backend/contrib/ethosu/tir_to_cs_translator.py | 76 +++-
python/tvm/relay/backend/contrib/ethosu/util.py | 15 +
python/tvm/relay/backend/vm.py | 17 +
python/tvm/relay/build_module.py | 20 +-
python/tvm/relay/frontend/caffe.py | 7 +-
python/tvm/relay/frontend/keras.py | 7 +-
python/tvm/relay/frontend/mxnet.py | 46 ++-
python/tvm/relay/frontend/onnx.py | 3 +-
python/tvm/relay/frontend/pytorch.py | 11 +-
python/tvm/relay/frontend/qnn_torch.py | 8 +-
python/tvm/relay/frontend/tensorflow_ops.py | 5 +-
python/tvm/relay/frontend/tflite.py | 39 +-
python/tvm/relay/op/contrib/ethosu.py | 352 ++++++++++++++++-
python/tvm/relay/op/nn/nn.py | 2 +-
python/tvm/relay/op/strategy/cuda.py | 2 +-
python/tvm/relay/op/strategy/generic.py | 26 +-
python/tvm/relay/op/strategy/x86.py | 18 +-
python/tvm/relay/qnn/op/layout_conversions.py | 2 +-
python/tvm/relay/qnn/op/qnn.py | 8 +-
python/tvm/relay/testing/dcgan.py | 3 +-
python/tvm/topi/generic/nn.py | 17 +
python/tvm/topi/nn/conv2d_transpose.py | 209 ++++++++--
python/tvm/topi/testing/conv2d_transpose_python.py | 40 +-
rust/tvm/tests/basics/src/tvm_add.py | 2 +-
src/driver/driver_api.cc | 2 +-
src/relay/backend/aot_executor_codegen.cc | 77 ++--
src/relay/backend/build_module.cc | 2 +-
.../contrib/example_target_hooks/relay_to_tir.cc | 12 +-
src/relay/backend/graph_executor_codegen.cc | 89 +++--
src/relay/backend/graph_plan_memory.cc | 52 +--
src/relay/backend/interpreter.cc | 152 ++++----
src/relay/backend/te_compiler.cc | 169 ++++----
src/relay/backend/vm/compiler.cc | 3 +-
src/relay/op/call/call.cc | 116 ++++++
src/relay/op/call/call.h | 74 ++++
src/relay/op/contrib/ethosu/binary_elementwise.cc | 301 ++++++++++++++
src/relay/op/contrib/ethosu/common.cc | 18 +
src/relay/op/contrib/ethosu/common.h | 11 +
src/relay/op/contrib/ethosu/pooling.cc | 2 +-
src/relay/op/memory/device_copy.cc | 17 +
src/relay/op/nn/convolution.h | 25 +-
src/relay/op/vm/vm.h | 2 +-
src/relay/transforms/device_domains.cc | 33 +-
src/relay/transforms/memory_alloc.cc | 16 +-
src/target/compilation_config.cc | 48 +--
src/tir/transforms/lower_warp_memory.cc | 1 +
tests/python/contrib/test_bnns/infrastructure.py | 2 +-
.../contrib/test_bnns/test_onnx_topologies.py | 2 +-
tests/python/contrib/test_ethosu/infra.py | 53 +++
tests/python/contrib/test_ethosu/test_codegen.py | 252 +++++++++++-
tests/python/contrib/test_ethosu/test_legalize.py | 188 +++++++++
.../test_ethosu/test_replace_binary_elementwise.py | 335 ++++++++++++++++
.../test_ethosu/test_tir_to_cs_translator.py | 434 +++++++++++++++++++++
.../contrib/test_ethosu/test_type_inference.py | 116 ++++++
.../python/contrib/test_hexagon/infrastructure.py | 4 +-
.../contrib/test_vitis_ai/test_vitis_ai_codegen.py | 25 +-
tests/python/driver/tvmc/test_compiler.py | 2 +-
tests/python/frontend/caffe/test_forward.py | 3 +-
tests/python/frontend/onnx/test_forward.py | 3 +
tests/python/frontend/tensorflow2/common.py | 6 +-
tests/python/relay/aot/aot_test_utils.py | 3 +-
tests/python/relay/test_op_level2.py | 60 +--
tests/python/relay/test_op_qnn_conv2_transpose.py | 53 ++-
tests/python/relay/test_pass_convert_op_layout.py | 16 +-
.../python/test_topi_group_conv2d_transpose.py | 156 ++++++++
.../unittest/test_micro_model_library_format.py | 4 +-
.../python/unittest/test_target_codegen_hexagon.py | 2 +-
.../test_tir_transform_lower_warp_memory.py | 32 +-
vta/scripts/tune_conv2d_transpose.py | 3 +-
vta/scripts/tune_resnet.py | 11 +-
.../python/integration/test_benchmark_gemm.py | 7 +-
.../integration/test_benchmark_topi_conv2d.py | 10 +-
.../integration/test_benchmark_topi_dense.py | 10 +-
.../test_benchmark_topi_group_conv2d.py | 10 +-
vta/tests/python/unittest/test_vta_insn.py | 16 +-
vta/tutorials/autotvm/tune_alu_vta.py | 3 +-
vta/tutorials/frontend/deploy_classification.py | 4 +-
vta/tutorials/frontend/deploy_detection.py | 4 +-
vta/tutorials/matrix_multiply.py | 4 +-
vta/tutorials/optimize/convolution_opt.py | 4 +-
vta/tutorials/optimize/matrix_multiply_opt.py | 4 +-
vta/tutorials/vta_get_started.py | 4 +-
web/tests/python/webgpu_rpc_test.py | 7 +-
119 files changed, 4378 insertions(+), 631 deletions(-)
copy include/tvm/relay/attrs/{debug.h => call.h} (61%)
create mode 100644
python/tvm/relay/backend/contrib/ethosu/op/binary_elementwise.py
create mode 100644
python/tvm/relay/backend/contrib/ethosu/te/binary_elementwise.py
copy python/tvm/relay/backend/contrib/ethosu/tir/{pooling.py =>
binary_elementwise.py} (57%)
create mode 100644 src/relay/op/call/call.cc
create mode 100644 src/relay/op/call/call.h
create mode 100644 src/relay/op/contrib/ethosu/binary_elementwise.cc
create mode 100644
tests/python/contrib/test_ethosu/test_replace_binary_elementwise.py
create mode 100644 tests/python/topi/python/test_topi_group_conv2d_transpose.py
