apache / tvm

38 Pull requests merged by 26 people

• [microTVM] Remove Arduino aot code

  #8869 merged Sep 4, 2021

• [ONNX] Add OpSet 13 implementation for Hardmax

  #8924 merged Sep 4, 2021

• [Onnx] Turn off flaky nllloss test for now

  #8919 merged Sep 4, 2021

• [UnitTests][Ethos-N] Mark unit tests as requiring Ethos-N

  #8873 merged Sep 4, 2021

• [microTVM] Temporarily remove mps2_an521 from CI

  #8927 merged Sep 4, 2021

• [Hexagon] Add trivial conv2d operator to Hexagon relay strategy

  #8915 merged Sep 3, 2021

• Sanitize names of input tensors in interface header

  #8720 merged Sep 3, 2021

• Remove LoweredModule

  #8886 merged Sep 3, 2021

• [TensorIR][Minor] Allow Tuple/Array in TE lowering

  #8916 merged Sep 3, 2021

• [BUG] Shape Func of Split Op Error

  #8887 merged Sep 3, 2021

• [ONNX] Add index_put operator

  #8894 merged Sep 3, 2021

• [AutoScheduler] Propogate global autotvm state to PopenPool workers

  #8913 merged Sep 3, 2021

• [Community] Add @Hzfengsy as TIR/Auto Codeowner

  #8912 merged Sep 3, 2021

• [UnitTests][Contrib] Enable contrib tensorrt/coreml unit tests

  #8902 merged Sep 2, 2021

• [BUG] DataType Bug In SplitRel

  #8899 merged Sep 2, 2021

• Enable python debug runtime for exported network libraries

  #8793 merged Sep 2, 2021

• Set default value of p in LpPool as 2

  #8866 merged Sep 2, 2021

• [Community] @Hzfengsy -> Committer

  #8908 merged Sep 2, 2021

• Trivial uTVM -> microTVM "spelling" fix to align with branding.

  #8905 merged Sep 2, 2021

• [Vulkan][Topi] Parametrizing additional topi tests, marking vulkan failures

  #8904 merged Sep 2, 2021

• Move to new style issue template system

  #8898 merged Sep 2, 2021

• [Onnx] Support Negative Log Loss

  #8872 merged Sep 2, 2021

• [ROCm][TVMC] Add ROCm to the TVMC driver

  #8896 merged Sep 2, 2021

• [TVMScript] Fix error reporting on Store

  #8895 merged Sep 1, 2021

• [Docker] Re-enabled automatic --tty flag when running bash.

  #8861 merged Sep 1, 2021

• [microTVM][RVM] Add Arduino RVM

  #8748 merged Sep 1, 2021

• [1/10] CMSIS-NN graph partitioner for softmax

  #8653 merged Sep 1, 2021

• [TVMScript] Fix printing ForNode annotations

  #8891 merged Sep 1, 2021

• [CI] make pre-commit hooks to run on every push instead of every commit

  #8888 merged Aug 31, 2021

• [TensorIR][M2a] CacheRead/Write

  #8863 merged Aug 31, 2021

• Change target string to Target object in the TE compiler and interpreter

  #8835 merged Aug 31, 2021

• refactor optimize GEMM on CPU tutorial

  #8825 merged Aug 31, 2021

• [ONNX][TOPI] Support select_last_index for argmin/max

  #8816 merged Aug 31, 2021

• [Vulkan][Target] Added the driver name to the vulkan target string.

  #8882 merged Aug 31, 2021

• [RPC] Fix ios_rpc build

  #8864 merged Aug 31, 2021

• [Community] @manupa-arm -> Committer

  #8870 merged Aug 30, 2021

• [RISCV] Add support for llvm parameter -mabi (aka -target-abi)

  #8860 merged Aug 30, 2021

• [TIR] GetBlockReadWriteRegion

  #8875 merged Aug 29, 2021

20 Pull requests opened by 18 people

• Deploy the Pretrained Model on Jetson Nano

  #8878 opened Aug 29, 2021

• [Relay][Quantization] Per-Channel FQ2I

  #8883 opened Aug 30, 2021

• [microTVM] Zephyr: Fix option name in PROJECT_OPTIONS

  #8884 opened Aug 30, 2021

• [PROFILING] Profiling over RPC

  #8885 opened Aug 30, 2021

• [Lowering] Add TargetDevice composite data structure.

  #8892 opened Aug 31, 2021

• [Doc] Current passes

  #8893 opened Sep 1, 2021

• Add sse4/avx2 support for fast x86 int8 (vpmaddubsw/vpmaddwd/vpaddd)

  #8897 opened Sep 1, 2021

• [Documentation] Document rewrite_once option

  #8900 opened Sep 1, 2021

• [Onnx] Add SoftmaxCrossEntropyLoss

  #8906 opened Sep 1, 2021

• [ONNX] Add support for QLinearConcat contrib op

  #8907 opened Sep 1, 2021

• [Relay, TOPI] Make Softmax op fusible with elemwise ops

  #8909 opened Sep 2, 2021

• [Bugfix] Add a nullptr check to tir.Buffer to fix the illegal memory access

  #8910 opened Sep 2, 2021

• Add manupa-arm to CODEOWNERS

  #8911 opened Sep 2, 2021

• Run InferType in LowerTEPass and remove per_target_modules_ from the Interpreter

  #8914 opened Sep 2, 2021

• [Relay] Add a non-recursive LetNode VisitExpr_ for LabelOps Pass to avoid stack overflow

  #8917 opened Sep 3, 2021

• [Bugfix] Fix visit_attrs error if its function pointer is equal to nullptr

  #8920 opened Sep 3, 2021

• [Layout] Unify dense op input layout

  #8921 opened Sep 3, 2021

• Arm(R) Cortex(R)-M55 CPU and Arm(R) Ethos(TM)-U55 NPU Demo App

  #8922 opened Sep 3, 2021

• [ONNX] [Test] fix GRU modification (linear before reset = False) and reduce tolerance for RNN tests

  #8923 opened Sep 3, 2021

• Fix incorrect AOT Memory Planning

  #8926 opened Sep 3, 2021

3 Issues closed by 3 people

• [Bug] [Relay] Error when compiling ONNX with LpPool

  #7241 closed Sep 3, 2021

• [TEST][FLAKY] tests/python/contrib/test_cudnn.py

  #5667 closed Sep 1, 2021

• Compiling random operation in pytorch to random in relay

  #8881 closed Aug 31, 2021

10 Issues opened by 8 people

• [Bug] Tuning PyTorch model with AutoTVM on "arm_cpu" runs tuning for both "arm_cpu" and "x86"

  #8929 opened Sep 4, 2021

• [VOTE] Adopt New Code Review Guideline

  #8928 opened Sep 3, 2021

• [Bug]Got "ValueError: not enough values to unpack (expected 2, got 1)" when BatchMatMul using two same tensors

  #8925 opened Sep 3, 2021

• [FLAKY] test_onnx_nodes[llvm-test_nllloss_NCd1d2d3_none_no_weight_negative_ii]

  #8918 opened Sep 3, 2021

• [Tracking][Vulkan] Extending topi/unit tests to run on Vulkan

  #8903 opened Sep 1, 2021

• [Tracking][Contrib] Known failing unit tests

  #8901 opened Sep 1, 2021

• [Bug] Error when compiling a ONNX model with Squeeze operator

  #8890 opened Aug 31, 2021

• [Bug] Error when compiling a ONNX model with Split operator

  #8889 opened Aug 31, 2021

• [PYTORCH] [FRONTEND] Not handled data types

  #8880 opened Aug 30, 2021

• v0.7 document？

  #8879 opened Aug 30, 2021

34 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• [Runtime] Pipeline Executor Initial patch.

  #8702 commented on Sep 3, 2021 • 78 new comments

• [2/6] Arm(R) Ethos(TM)-U NPU Relay passes and Conv2D op

  #8795 commented on Sep 4, 2021 • 24 new comments

• [Tracking Issue][ONNX] Quantized operator support in ONNX importer

  #8838 commented on Sep 3, 2021 • 20 new comments

• [BYOC][TensorRT] Add TensorRT own int8 calibration support to TensorRT BYOC integration

  #8808 commented on Sep 3, 2021 • 16 new comments

• [Hexagon] Add support for linked-in model parameters

  #8865 commented on Sep 4, 2021 • 16 new comments

• [Draft PR] Visualization Relay IR on terminal

  #8668 commented on Sep 4, 2021 • 9 new comments

• [2/10] CMSIS-NN code generator for softmax

  #8833 commented on Sep 4, 2021 • 8 new comments

• [AutoTVM] Use popenpool in local_executor

  #8851 commented on Sep 2, 2021 • 6 new comments

• Possible flakey tests/micro/zephyr/test_zephyr.py "test_relay[mps2_an521]"

  #8728 commented on Sep 3, 2021 • 5 new comments

• [RELAY][FRONTEND] Initial OneFlow frontend support.

  #8790 commented on Sep 4, 2021 • 5 new comments

• Unit Test java GPU failed - java.io.IOException: java.lang.RuntimeException: Failed to serialize

  #8876 commented on Sep 3, 2021 • 4 new comments

• [OpenCL] Add vectorization to cuda conv2d_nhwc schedule

  #8636 commented on Sep 4, 2021 • 4 new comments

• [WIP][Runtime]Pipeline Executor For Compute graph pipeline

  #7892 commented on Sep 3, 2021 • 3 new comments

• [Frontend][ONNX] Add onnx support for `LessOrEqual`, `GreaterOrEqual`

  #8626 commented on Aug 30, 2021 • 2 new comments

• Frontend: add onnx GlobalLpPool op

  #8845 commented on Sep 2, 2021 • 2 new comments

• NotImplementedError: The following operators are not implemented: ['aten::stft']

  #8877 commented on Aug 31, 2021 • 1 new comment

• Fusion of operations and cast in mobilenet v1 conv2d causing large feature maps

  #8717 commented on Sep 1, 2021 • 1 new comment

• [Frontend][ONNX] No support for `RandomUniform`

  #8632 commented on Sep 2, 2021 • 1 new comment

• [Bug] Error when compiling a ONNX model with Gemm operator

  #7262 commented on Sep 2, 2021 • 1 new comment

• [RFC][Tracking Issue] TensorIR Scheduling

  #7527 commented on Sep 3, 2021 • 1 new comment

• [RFC][Tracking Issue][AMP] Tracking Issue for Mixed Precision Pass

  #8296 commented on Sep 3, 2021 • 1 new comment

• [BUG]Import the DLRM NET graph from pytorch to Relay failed

  #8374 commented on Sep 3, 2021 • 1 new comment

• [Relay] Make more passes iterative on a-normal graphs.

  #8434 commented on Sep 1, 2021 • 1 new comment

• Pass resource_handle to operators with unpacked API

  #8452 commented on Sep 4, 2021 • 1 new comment

• Add group_conv2d_transpose_nchw to support `groups` argument

  #8799 commented on Aug 31, 2021 • 1 new comment

• [CI][VitisAI] Update CI Vitis AI PyXIR version to v0.3.1

  #8814 commented on Sep 3, 2021 • 1 new comment

• Error emitting Optimized C Code

  #8783 commented on Aug 31, 2021 • 0 new comments

• [RFC][Tracking Issue] CMSIS-NN Integration

  #8646 commented on Sep 1, 2021 • 0 new comments

• pooling convert layout bug in pooling.cc and in test_pass_convert_op_layout.py? (our resolutions are provided)

  #8848 commented on Sep 2, 2021 • 0 new comments

• [ROADMAP] TVM v0.8 Roadmap

  #7434 commented on Sep 2, 2021 • 0 new comments

• [TIR][VM] Revert a change to lower_tvm_builtin.cc from #6126

  #8274 commented on Sep 3, 2021 • 0 new comments

• [ETHOSN] Add support for non-default ethos-n78 configurations

  #8300 commented on Sep 1, 2021 • 0 new comments

• [PyTorch][WIP]Add PyTorchTVM: compile torchscript to tvm and export as pytorch_op

  #8777 commented on Sep 3, 2021 • 0 new comments

• DRAFT [Relay] Remove DeviceMap from LowerTE

  #8788 commented on Sep 4, 2021 • 0 new comments