I am try tuning a tensorflow model on x86, when apply graph tuning, I have a trouble like following,
2019-11-11 17:28:36,479 INFO Start to benchmark layout transformation...
2019-11-11 17:28:36,479 INFO Benchmarking layout transformation successful.
2019-11-11 17:28:36,480 INFO Start to run PBQP algorithm...
Traceback (most recent call last):
File "./tvm_complie_bert.py", line 181, in <module>
main()
File "./tvm_complie_bert.py", line 177, in main
tune(mod, lib, params)
File "./tvm_complie_bert.py", line 123, in tune
executor.run()
File "/home/kim/repo/tf1.15.0/lib64/python3.6/site-packages/tvm-0.6.dev0-py3.6-linux-x86_64.egg/tvm/autotvm/graph_tuner/pbqp_tuner.py", line 292, in run
self._forward()
File "/home/kim/repo/tf1.15.0/lib64/python3.6/site-packages/tvm-0.6.dev0-py3.6-linux-x86_64.egg/tvm/autotvm/graph_tuner/pbqp_tuner.py", line 200, in _forward
if self._buckets[1]:
IndexError: list index out of range