Hi all,
I came across an unexpected crash when executing the flowing simple script.
It threw the a wired crash message: "Tensor[(5), float32] does not match Tensor[(5), float64] "
If I replace dtype=float64 with dtype=float32 in this script, It can run well.
It seems float64 is a supported data type. If this given script is correct? or trigger an inner bug of TVM?
Script
import torch
from tvm import relay
m = torch.nn.BatchNorm1d(5)
input_data = [torch.randn([0, 5, 9], dtype=torch.float64)]
trace = torch.jit.trace(m, input_data)
input_shapes = [('input0', ([0, 5, 9],'float64') )]
mod, params = relay.frontend.from_pytorch(trace, input_shapes)
Full traceback
Traceback (most recent call last):
File "test.py", line 11, in <module>
mod, params = relay.frontend.from_pytorch(trace, input_shapes)
File "/workplace/software/tvm/tvm_/python/tvm/relay/frontend/pytorch.py", line 5002, in from_pytorch
outputs = converter.convert_operators(operator_nodes, outputs, ret_name)
File "/workplace/software/tvm/tvm_/python/tvm/relay/frontend/pytorch.py", line 4263, in convert_operators
self.record_output_type(relay_out)
File "/workplace/software/tvm/tvm_/python/tvm/relay/frontend/pytorch.py", line 238, in record_output_type
self.infer_type_with_prelude(output)
File "/workplace/software/tvm/tvm_/python/tvm/relay/frontend/pytorch.py", line 174, in infer_type_with_prelude
body = self.infer_type(val, self.prelude.mod)
File "/workplace/software/tvm/tvm_/python/tvm/relay/frontend/pytorch.py", line 167, in infer_type
new_mod = transform.InferType()(new_mod)
File "/workplace/software/tvm/tvm_/python/tvm/ir/transform.py", line 160, in __call__
return _ffi_transform_api.RunPass(self, mod)
File "/workplace/software/tvm/tvm_/python/tvm/_ffi/_ctypes/packed_func.py", line 237, in __call__
raise get_last_ffi_error()
tvm._ffi.base.TVMError: Traceback (most recent call last):
10: TVMFuncCall
9: tvm::runtime::PackedFuncObj::Extractor<tvm::runtime::PackedFuncSubObj<tvm::runtime::TypedPackedFunc<tvm::IRModule (tvm::transform::Pass, tvm::IRModule)>::AssignTypedLambda<tvm::transform::$_6>(tvm::transform::$_6, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >)::{lambda(tvm::runtime::TVMArgs const&, tvm::runtime::TVMRetValue*)#1}> >::Call(tvm::runtime::PackedFuncObj const*, tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*)
8: tvm::transform::Pass::operator()(tvm::IRModule) const
7: tvm::transform::Pass::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
6: tvm::transform::ModulePassNode::operator()(tvm::IRModule, tvm::transform::PassContext const&) const
5: tvm::runtime::PackedFuncObj::Extractor<tvm::runtime::PackedFuncSubObj<tvm::runtime::TypedPackedFunc<tvm::IRModule (tvm::IRModule, tvm::transform::PassContext)>::AssignTypedLambda<tvm::relay::transform::InferType()::$_2>(tvm::relay::transform::InferType()::$_2)::{lambda(tvm::runtime::TVMArgs const&, tvm::runtime::TVMRetValue*)#1}> >::Call(tvm::runtime::PackedFuncObj const*, tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*)
4: tvm::DiagnosticContext::Render()
3: tvm::DiagnosticRenderer::Render(tvm::DiagnosticContext const&)
2: tvm::runtime::PackedFuncObj::Extractor<tvm::runtime::PackedFuncSubObj<tvm::runtime::TypedPackedFunc<void (tvm::DiagnosticContext)>::AssignTypedLambda<tvm::TerminalRenderer(std::ostream&)::$_10>(tvm::TerminalRenderer(std::ostream&)::$_10)::{lambda(tvm::runtime::TVMArgs const&, tvm::runtime::TVMRetValue*)#1}> >::Call(tvm::runtime::PackedFuncObj const*, tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*)
1: tvm::ReportAt(tvm::DiagnosticContext const&, std::ostream&, tvm::Span const&, tvm::Diagnostic const&)
0: _ZN3tvm7runtime6detail
File "/workplace/software/tvm/tvm_/src/ir/diagnostic.cc", line 264
TVMError: The source maps are not populated for this module. Please use `tvm.relay.transform.AnnotateSpans` to attach source maps for error reporting.
Error: The Relay type checker is unable to show the following types match.
In particular `Tensor[(5), float32]` does not match `Tensor[(5), float64]`