schliffen opened a new issue #5449:
URL: https://github.com/apache/incubator-tvm/issues/5449


   We are trying to convert a mxnet model into tvm (v7) in windows using the 
following part of code:
   loading model (works fine):
    `      mx_sym, args, auxs = mx.model.load_checkpoint(root + ap.mdlrt, 0)
           mod, params = relay.frontend.from_mxnet(mx_sym, shape_dict, dtype, 
arg_params=args, aux_params=auxs)
   `
   build tvm model (fails):
   `    with relay.build_config(opt_level=3):
           graph, mod, params = relay.build(mod, target, params=params)
   `
   in this relay.build we faced with the following error (full log):
   `[08:31:15] 
C:\Jenkins\workspace\mxnet-tag\mxnet\src\nnvm\legacy_json_util.cc:209: Loading 
symbol saved by previous version v1.2.0. Attempting to upgrade...
   [08:31:15] 
C:\Jenkins\workspace\mxnet-tag\mxnet\src\nnvm\legacy_json_util.cc:217: Symbol 
successfully upgraded!
   Traceback (most recent call last):
     File "E:/TVM/faceTransfer/mxnet2Tvm.py", line 96, in save_tvm_model
       graph, mod, params = relay.build(mod, target, params=params)
     File 
"C:\Python\Python37\site-packages\tvm-0.7.dev1-py3.7-win-amd64.egg\tvm\relay\build_module.py",
 line 251, in build
       graph_json, mod, params = bld_mod.build(mod, target, target_host, params)
     File 
"C:\Python\Python37\site-packages\tvm-0.7.dev1-py3.7-win-amd64.egg\tvm\relay\build_module.py",
 line 120, in build
       self._build(mod, target, target_host)
     File 
"C:\Python\Python37\site-packages\tvm-0.7.dev1-py3.7-win-amd64.egg\tvm\_ffi\_ctypes\packed_func.py",
 line 215, in __call__
       ctypes.byref(ret_val), ctypes.byref(ret_tcode)) != 0:
   OSError: exception: access violation reading 0x0000000000000000`
   
   system details: Windows 10, Visual Studio-2015, c++14, cuda 10.1, llvm 
6.0.1, tvm v7


----------------------------------------------------------------
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


Reply via email to