Hello,
I a have been following a tutorial to compile mxnet model: https://tvm.apache.org/docs/how_to/compile_models/from_mxnet.html
But I got following error:
/Users/xxxx/Documents/workspace/xxxx/mxnet/src/storage/storage.cc:202: Using Pooled (Naive) StorageManager for CPU
x (1, 3, 224, 224)
Traceback (most recent call last):
File "/Users/xxxx/Documents/workspace/playground/tvm-playground/mxnettest.py", line 45, in <module>
mod, params = relay.frontend.from_mxnet(block, shape_dict)
File "/Users/xxxx/Documents/workspace/playground/tvm-playground/venv/lib/python3.9/site-packages/tvm-0.9.dev1087+gb99a09682-py3.9-macosx-12-arm64.egg/tvm/relay/frontend/mxnet.py", line 2921, in from_mxnet
sym = symbol(*inputs)
File "/Users/xxxx/Documents/workspace/xxxx/mxnet/python/mxnet/util.py", line 304, in _with_np_shape
return func(*args, **kwargs)
File "/Users/xxxx/Documents/workspace/xxxx/mxnet/python/mxnet/util.py", line 486, in _with_np_array
return func(*args, **kwargs)
File "/Users/xxxx/Documents/workspace/xxxx/mxnet/python/mxnet/gluon/block.py", line 1588, in __call__
_check_block_input_np_ndarrays([x, *args])
File "/Users/xxxx/Documents/workspace/xxxx/mxnet/python/mxnet/gluon/utils.py", line 505, in _check_block_input_np_ndarrays
_check_block_input_np_ndarrays(i)
File "/Users/xxxx/Documents/workspace/xxxx/mxnet/python/mxnet/gluon/utils.py", line 500, in _check_block_input_np_ndarrays
raise TypeError("Block's inputs must be of type `mxnet.numpy.ndarray`, "
TypeError: Block's inputs must be of type `mxnet.numpy.ndarray`, while got output type <class 'mxnet.symbol.symbol.Symbol'>
I am using mxnet==2.0.0 and tvm-0.9.dev1087+gb99a09682.