Airtnp opened a new issue #4717: [AutoTVM] Failed to run autotvm example on GEMM URL: https://github.com/apache/incubator-tvm/issues/4717 I was trying to use code from https://zhuanlan.zhihu.com/p/75203171 (it's a Chinese post, but it's essential the same as [Writing tunable template and Using auto-tuner](https://docs.tvm.ai/tutorials/autotvm/tune_simple_template.html)) with modified size 4096^3. However, in my AWS instance, it keeps reporting tons of similar error like this > DEBUG:autotvm:No: 199 GFLOPS: 0.00/0.00 result: MeasureResult(costs=(TVMError('Traceback (most recent call last):\n [bt] (3) /home/ubuntu/.local/lib/python3.6/site-packages/tvm-0.6.0-py3.6-linux-x86_64.egg/tvm/libtvm.so(TVMFuncCall+0x61) [0x7f51e3d495e1]\n [bt] (2) /home/ubuntu/.local/lib/python3.6/site-packages/tvm-0.6.0-py3.6-linux-x86_64.egg/tvm/libtvm.so(_ZNSt17_Function_handlerIFvN3tvm7runtime7TVMArgsEPNS1_11TVMRetValueEEZNS1_15TypedPackedFuncIFRNS0_5StageES7_NS0_7IterVarEEE17AssignTypedLambdaIZNS1_8Registry15set_body_methodIS7_S8_IS9_EEERSD_MT_FT0_DpT1_EEUlS7_S9_E_EEvSG_EUlRKS2_S4_E_E9_M_invokeERKSt9_Any_dataOS2_OS4_+0x98) [0x7f51e3570c48]\n [bt] (1) /home/ubuntu/.local/lib/python3.6/site-packages/tvm-0.6.0-py3.6-linux-x86_64.egg/tvm/libtvm.so(tvm::Stage::vectorize(tvm::IterVar)+0xb0) [0x7f51e38e7c70]\n [bt] (0) /home/ubuntu/.local/lib/python3.6/site-packages/tvm-0.6.0-py3.6-linux-x86_64.egg/tvm/libtvm.so(dmlc::LogMessageFatal::~LogMessageFatal()+0x32) [0x7f51e35375e2]\n File "/home/ubuntu/tvm/src/schedule/schedule_lang.cc", line 377\nTVMError: Check failed: var->iter_type == kDataPar || var->iter_type == kOpaque || var->iter_type == kUnrolled || var->iter_type == kVectorized || var->iter_type == kTensorized || var->iter_type == kParallelized: Cannot vectorize on CommReduce',),), error_no=2, all_cost=0.006806135177612305, timestamp=1579123976.0153828) [('tile_x', [-1, 8, 32]), ('tile_y', [-1, 16, 32]), ('tile_k', [-1, 2048]), ('reorder', (2, 0, 1)), ('ann', ['unroll', 'vec', 'unroll'])],,None,10215900 I tried with code in [Writing tunable template and Using auto-tuner](https://docs.tvm.ai/tutorials/autotvm/tune_simple_template.html) and it also reports errors like > No: 9 GFLOPS: 0.00/0.00 result: MeasureResult(costs=(RuntimeError('Traceback (most recent call last):\n [bt] (3) /home/ubuntu/.local/lib/python3.6/site-packages/tvm-0.6.0-py3.6-linux-x86_64.egg/tvm/libtvm.so(TVMFuncCall+0x61) [0x7f6bdfd335e1]\n [bt] (2) /home/ubuntu/.local/lib/python3.6/site-packages/tvm-0.6.0-py3.6-linux-x86_64.egg/tvm/libtvm.so(std::_Function_handler<void (tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*), tvm::runtime::RPCModuleNode::WrapRemote(void*)::{lambda(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*)#1}>::_M_invoke(std::_Any_data const&, tvm::runtime::TVMArgs&&, tvm::runtime::TVMRetValue*&&)+0x41) [0x7f6bdfd914f1]\n [bt] (1) /home/ubuntu/.local/lib/python3.6/site-packages/tvm-0.6.0-py3.6-linux-x86_64.egg/tvm/libtvm.so(tvm::runtime::RPCSession::CallFunc(void*, tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*, void* (*)(int, tvm::runtime::TVMArgValue const&), tvm::runtime::PackedFunc const*)+0x167) [0x7f6bdfd9afb7]\n [bt] (0) /home/ubuntu/.local/lib/python3.6/site-packages/tvm-0.6.0-py3.6-linux-x86_64.egg/tvm/libtvm.so(dmlc::LogMessageFatal::~Lo',),), error_no=4, all_cost=10.207964420318604, timestamp=1579122195.5362496) [('tile_y', [-1, 128]), ('tile_x', [-1, 8])],,None,46 What's wrong with the code? Thank you for your help!
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
