You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
[RELAY][PASS] AlterOpLayout handling for Tuple #2382
File "./tests/python/frontend/tensorflow/test_forward.py", line 1060, in <module>
test_forward_inception_v3()
File "./tests/python/frontend/tensorflow/test_forward.py", line 701, in test_forward_inception_v3
tvm_output = run_tvm_graph(graph_def, data, 'input')
File "./tests/python/frontend/tensorflow/test_forward.py", line 60, in run_tvm_graph
graph, lib, params = relay.build(sym, target, params=params)
File "/home/srk/.local/lib/python3.6/site-packages/tvm-0.5.dev0-py3.6-linux-x86_64.egg/tvm/relay/build_module.py", line 227, in build
func = ir_pass.infer_type(func)
File "/home/srk/.local/lib/python3.6/site-packages/tvm-0.5.dev0-py3.6-linux-x86_64.egg/tvm/relay/ir_pass.py", line 44, in infer_type
return _ir_pass.infer_type(expr, mod)
File "/home/srk/.local/lib/python3.6/site-packages/tvm-0.5.dev0-py3.6-linux-x86_64.egg/tvm/_ffi/_ctypes/function.py", line 185, in __call__
ctypes.byref(ret_val), ctypes.byref(ret_tcode)))
File "/home/srk/.local/lib/python3.6/site-packages/tvm-0.5.dev0-py3.6-linux-x86_64.egg/tvm/_ffi/base.py", line 72, in check_call
raise TVMError(py_str(_LIB.TVMGetLastError()))
tvm._ffi.base.TVMError: [16:21:13] /home/srk/work/DMLC/tvm/src/relay/pass/type_infer.cc:317: the function is provided too many arguments (nullptr)
Tuple arguments not retained as Tuple after layout handling for operators like Concatenate.
Activity
srkreddy1238 commentedon Jan 7, 2019
#2383
srkreddy1238 commentedon Jan 7, 2019
#2536 fix this issue.