[TOPI] Error while doing softmax on concatenated array #1024
Description
Traceback (most recent call last):
File "topi/tests/python_cpp/test_topi_region.py", line 50, in <module>
test_region()
File "topi/tests/python_cpp/test_topi_region.py", line 47, in test_region
verify_region(1, 2, 8, 2, 2, 1, 1, 1)
File "topi/tests/python_cpp/test_topi_region.py", line 43, in verify_region
check_device(device)
File "topi/tests/python_cpp/test_topi_region.py", line 38, in check_device
func = tvm.build(s, [A, B], device)
File "/home/pariksheet/.local/lib/python3.6/site-packages/tvm-0.2.0-py3.6-linux-x86_64.egg/tvm/build_module.py", line 467, in build
mhost = codegen.build_module(fhost, str(target_host))
File "/home/pariksheet/.local/lib/python3.6/site-packages/tvm-0.2.0-py3.6-linux-x86_64.egg/tvm/codegen.py", line 20, in build_module
return _Build(lowered_func, target)
File "/home/pariksheet/.local/lib/python3.6/site-packages/tvm-0.2.0-py3.6-linux-x86_64.egg/tvm/_ffi/function.py", line 280, in my_api_func
return flocal(*args)
File "tvm/_ffi/_cython/./function.pxi", line 264, in core.FunctionBase.__call__
File "tvm/_ffi/_cython/./function.pxi", line 213, in core.FuncCall
File "tvm/_ffi/_cython/./function.pxi", line 205, in core.FuncCall3
File "tvm/_ffi/_cython/./base.pxi", line 131, in core.CALL
tvm._ffi.base.TVMError: b'[17:39:01] src/codegen/llvm/codegen_llvm.cc:632: unknown intrinsic indeterminate_expression\n\nStack trace returned 10 entries:\n[bt] (0) /home/pariksheet/.local/lib/python3.6/site-packages/tvm-0.2.0-py3.6-linux-x86_64.egg/tvm/libtvm.so(dmlc::StackTrace[abi:cxx11]()+0x53) [0x7f74bd130003]\n[bt] (1) /home/pariksheet/.local/lib/python3.6/site-packages/tvm-0.2.0-py3.6-linux-x86_64.egg/tvm/libtvm.so(dmlc::LogMessageFatal::~LogMessageFatal()+0x39) [0x7f74bd130879]\n[bt] (2) /home/pariksheet/.local/lib/python3.6/site-packages/tvm-0.2.0-py3.6-linux-x86_64.egg/tvm/libtvm.so(tvm::codegen::CodeGenLLVM::CreateIntrinsic(HalideIR::Internal::Call const*)+0x8b) [0x7f74bd3ada8b]\n[bt] (3) /home/pariksheet/.local/lib/python3.6/site-packages/tvm-0.2.0-py3.6-linux-x86_64.egg/tvm/libtvm.so(tvm::codegen::CodeGenCPU::CreateIntrinsic(HalideIR::Internal::Call const*)+0x40) [0x7f74bd3ba110]\n[bt] (4) /home/pariksheet/.local/lib/python3.6/site-packages/tvm-0.2.0-py3.6-linux-x86_64.egg/tvm/libtvm.so(tvm::codegen::CodeGenLLVM::VisitExpr_(HalideIR::Internal::Call const*)+0x61) [0x7f74bd3a51f1]\n[bt] (5) /home/pariksheet/.local/lib/python3.6/site-packages/tvm-0.2.0-py3.6-linux-x86_64.egg/tvm/libtvm.so(tvm::IRFunctor<llvm::Value* (tvm::NodeRef const&, tvm::ir::ExprFunctor<llvm::Value* (HalideIR::Expr const&)>*)>::operator()(tvm::NodeRef const&, tvm::ir::ExprFunctor<llvm::Value* (HalideIR::Expr const&)>*) const+0x7d) [0x7f74bd38e61d]\n[bt] (6) /home/pariksheet/.local/lib/python3.6/site-packages/tvm-0.2.0-py3.6-linux-x86_64.egg/tvm/libtvm.so(tvm::codegen::CodeGenLLVM::VisitExpr_(HalideIR::Internal::Load const*)+0x75) [0x7f74bd3a8f75]\n[bt] (7) /home/pariksheet/.local/lib/python3.6/site-packages/tvm-0.2.0-py3.6-linux-x86_64.egg/tvm/libtvm.so(tvm::IRFunctor<llvm::Value* (tvm::NodeRef const&, tvm::ir::ExprFunctor<llvm::Value* (HalideIR::Expr const&)>*)>::operator()(tvm::NodeRef const&, tvm::ir::ExprFunctor<llvm::Value* (HalideIR::Expr const&)>*) const+0x7d) [0x7f74bd38e61d]\n[bt] (8) /home/pariksheet/.local/lib/python3.6/site-packages/tvm-0.2.0-py3.6-linux-x86_64.egg/tvm/libtvm.so(tvm::codegen::CodeGenLLVM::VisitStmt_(HalideIR::Internal::Store const*)+0x108) [0x7f74bd3a9698]\n[bt] (9) /home/pariksheet/.local/lib/python3.6/site-packages/tvm-0.2.0-py3.6-linux-x86_64.egg/tvm/libtvm.so(tvm::IRFunctor<void (tvm::NodeRef const&, tvm::ir::StmtFunctor<void (HalideIR::Internal::Stmt const&)>*)>::operator()(tvm::NodeRef const&, tvm::ir::StmtFunctor<void (HalideIR::Internal::Stmt const&)>*) const+0x11d) [0x7f74bd30443d]\n\n'
This the error i am getting while softmax an tensor array which is been concatenated. Can anyone help?
Code snippet:
auto x = concatenate(softmax_input, 2);
Tensor softmax_output = softmax(x, 2);
Shape of the axis along which the array is concatenated is an expression. If i manually change the shape to an integer value, then it is working fine.