I am stuck with this error. I have a torch model and wanted to test the performance of TVM on this torch model. For this I converted the torch model in onnx format and then followed same procedure as this repo suggested. But it is giving me error
tvm_out = executor.evaluate()(tvm.nd.array(data.astype(dtype)), **params)
File "/.local/lib/python3.6/site-packages/tvm-0.8.dev113+g3d624ec-py3.6-linux-x86_64.egg/tvm/relay/backend/interpreter.py", line 172, in evaluate
return self._make_executor()
File "/.local/lib/python3.6/site-packages/tvm-0.8.dev113+g3d624ec-py3.6-linux-x86_64.egg/tvm/relay/build_module.py", line 370, in _make_executor
raise ValueError("Graph Runtime only supports static graphs, got output type", ret_type)
ValueError: ('Graph Runtime only supports static graphs, got output type', TensorType([?, 1], float32)
Exact same code is running on different computer with lower version of TVM
lemon234071
commented
3 years ago
I am stuck with this error. I have a torch model and wanted to test the performance of TVM on this torch model. For this I converted the torch model in onnx format and then followed same procedure as this repo suggested. But it is giving me error
tvm_out = executor.evaluate()(tvm.nd.array(data.astype(dtype)), **params) File "/.local/lib/python3.6/site-packages/tvm-0.8.dev113+g3d624ec-py3.6-linux-x86_64.egg/tvm/relay/backend/interpreter.py", line 172, in evaluate return self._make_executor() File "/.local/lib/python3.6/site-packages/tvm-0.8.dev113+g3d624ec-py3.6-linux-x86_64.egg/tvm/relay/build_module.py", line 370, in _make_executor raise ValueError("Graph Runtime only supports static graphs, got output type", ret_type) ValueError: ('Graph Runtime only supports static graphs, got output type', TensorType([?, 1], float32)
Exact same code is running on different computer with lower version of TVM