This is an automated email from the ASF dual-hosted git repository. mbrookhart pushed a change to branch ci-docker-staging in repository https://gitbox.apache.org/repos/asf/tvm.git.
discard 13e27ef Merge branch 'update_onnx' of github.com:mbrookhart/tvm into update_onnx discard 8dd9088 next try at docker images discard 92303bb skip a test until update complete discard 7955b21 try updating docker images again discard 0a3c5ca disable test until CI update complete discard aec720b Don't force output shape for conv transpose tests, add 1D and 3D cases discard 4bdb558 support convtranspose opset 11 autopadding discard 4e934d9 point jenkins at new docker discard 6bf7fd0 add failing onnx tets add 073becd [MyPy] Minimal type checking on TIR schedule (#8367) add 1d64fa5 Update the tvmc tutorial with additional requirements (#8334) add 39e1ffe Support QLinearAdd from onnx runtime com.microsoft contrib ops. (#8305) add c989e4a [Metal] Add pass for splitting kernel with huge number of args (#8313) add 578f617 [Tuning] Allow multiprocessing spawn to work (on macOS llvm at least) (#8363) add 7504a9c fix ci-arm build process (#8377) add a66186b [FIX] Fix depthwise conv2d on non-cuda GPU platforms (#8379) add 2e47947 [cuDNN] Add support for log_softmax (#8369) add 9112b6e Allow tvmc to compile models with AOT executor in MLF (#8331) add 29e958d [TIR][TVMScript] specialize (#8354) add 6f600f1 [Refactor] Remove dead code from depthwise_conv2d for Intel graphics (#8381) add ab01abc [BugFix][Relay] Fix type relation for batch_matmul (#8376) add 22204be fix keras install (#8391) add 7b898d0 Fix np.int and np.float usage in the tree. (#8389) add 354d996 Add missing annotation for requires_gpu in test_topi_dense.py Requires GPU (#8387) add 970aeff Add "operator" style to Model Library Format (#8072) add 2e3d617 macOS is now supported (#8396) add 7e3f068 [microTVM] Add Nucleo stm32l4r5zi board to zephyr (#8386) add e19e979 [Torch] Remove unused conversion (#8397) add e32d47e [Arith] Inverse affine map (#8384) add a00d211 Actually add Compute Library tests to the Jenkins File (#8394) add bf1b4be add failing onnx tets add 8d20c3e point jenkins at new docker add c92cec0 support convtranspose opset 11 autopadding add 1e4b630 Don't force output shape for conv transpose tests, add 1D and 3D cases add dcdfe62 disable test until CI update complete add 2cbe52d try updating docker images again add 4564df5 skip a test until update complete add c41a4eb next try at docker images This update added new revisions after undoing existing revisions. That is to say, some revisions that were in the old version of the branch are not in the new version. This situation occurs when a user --force pushes a change and generates a repository containing something like this: * -- * -- B -- O -- O -- O (13e27ef) \ N -- N -- N refs/heads/ci-docker-staging (c41a4eb) You should already have received notification emails for all of the O revisions, and so the following emails describe only the N revisions from the common base, B. Any revisions marked "omit" are not gone; other references still refer to them. Any revisions marked "discard" are gone forever. No new revisions were added by this update. Summary of changes: Jenkinsfile | 1 + ...40dk_nrf5340_cpuapp.conf => nucleo_l4r5zi.conf} | 4 +- apps/microtvm/zephyr/aot_demo/src/zephyr_uart.c | 6 + .../{nucleo_f746zg.conf => nucleo_l4r5zi.conf} | 3 +- .../ubuntu_download_arm_compute_lib_binaries.sh | 2 +- docker/install/ubuntu_install_tensorflow.sh | 2 +- include/tvm/arith/iter_affine_map.h | 21 ++ include/tvm/relay/transform.h | 7 + include/tvm/tir/analysis.h | 2 +- include/tvm/tir/buffer.h | 1 + include/tvm/tir/function.h | 38 +++ mypy.ini | 20 +- python/tvm/arith/__init__.py | 7 +- python/tvm/arith/iter_affine_map.py | 27 ++ python/tvm/autotvm/measure/measure_methods.py | 63 ++-- python/tvm/autotvm/task/task.py | 2 +- python/tvm/contrib/cudnn.py | 26 ++ python/tvm/contrib/utils.py | 13 + python/tvm/driver/build_module.py | 29 +- python/tvm/driver/tvmc/compiler.py | 6 +- python/tvm/driver/tvmc/model.py | 13 +- python/tvm/micro/contrib/zephyr.py | 1 + python/tvm/micro/model_library_format.py | 208 +++++++++++-- python/tvm/relay/build_module.py | 26 +- python/tvm/relay/frontend/onnx.py | 44 ++- python/tvm/relay/frontend/pytorch.py | 3 - python/tvm/relay/op/nn/_nn.py | 2 +- python/tvm/relay/op/strategy/cuda.py | 22 +- python/tvm/relay/op/strategy/generic.py | 15 +- python/tvm/relay/op/strategy/hls.py | 15 +- python/tvm/relay/op/strategy/x86.py | 15 +- python/tvm/relay/transform/transform.py | 11 + python/tvm/target/target.py | 16 +- python/tvm/tir/function.py | 55 +++- python/tvm/tir/schedule/block_scope.py | 12 +- python/tvm/tir/schedule/schedule.py | 28 +- python/tvm/tir/schedule/state.py | 10 +- python/tvm/topi/cuda/depthwise_conv2d.py | 10 +- python/tvm/topi/cuda/softmax.py | 10 + python/tvm/topi/intel_graphics/depthwise_conv2d.py | 183 ----------- python/tvm/topi/nn/softmax.py | 5 +- python/tvm/topi/testing/depthwise_conv2d_python.py | 36 +-- src/arith/iter_affine_map.cc | 142 +++++++++ src/printer/model_library_format_printer.cc | 81 +++++ src/printer/text_printer.h | 10 + src/printer/tir_text_printer.cc | 10 + src/relay/backend/build_module.cc | 6 + src/relay/op/nn/nn.cc | 3 +- src/relay/transforms/split_args.cc | 95 ++++++ src/runtime/contrib/cudnn/softmax.cc | 91 +++--- src/target/source/codegen_metal.cc | 9 +- src/target/source/codegen_metal.h | 3 +- src/target/target.cc | 3 - src/target/target_kind.cc | 5 + src/tir/ir/specialize.cc | 337 +++++++++++++++++++++ tests/lint/check_file_type.py | 2 + tests/micro/zephyr/conftest.py | 7 +- tests/python/contrib/test_cudnn.py | 28 +- tests/python/driver/tvmc/conftest.py | 60 +--- tests/python/driver/tvmc/test_mlf.py | 47 ++- tests/python/driver/tvmc/test_runner.py | 5 +- tests/python/frontend/onnx/test_forward.py | 62 +++- tests/python/integration/test_tuning.py | 107 ++++--- tests/python/relay/test_op_level10.py | 37 ++- tests/python/relay/test_pass_split_args.py | 96 ++++++ tests/python/topi/python/test_topi_dense.py | 1 + .../topi/python/test_topi_depthwise_conv2d.py | 121 ++++---- .../python/unittest/test_arith_iter_affine_map.py | 61 ++++ .../unittest/test_micro_model_library_format.py | 68 ++++- tests/python/unittest/test_target_target.py | 22 +- tests/python/unittest/test_tir_specialize.py | 199 ++++++++++++ tests/scripts/task_mypy.sh | 8 +- tutorials/get_started/tvmc_command_line_driver.py | 25 +- 73 files changed, 2155 insertions(+), 616 deletions(-) copy apps/microtvm/zephyr/aot_demo/boards/{nrf5340dk_nrf5340_cpuapp.conf => nucleo_l4r5zi.conf} (92%) copy apps/microtvm/zephyr/host_driven/boards/{nucleo_f746zg.conf => nucleo_l4r5zi.conf} (92%) create mode 100644 src/printer/model_library_format_printer.cc create mode 100644 src/relay/transforms/split_args.cc create mode 100644 src/tir/ir/specialize.cc create mode 100644 tests/python/relay/test_pass_split_args.py create mode 100644 tests/python/unittest/test_tir_specialize.py