Cannot find config for target=cuda for the dense layer


#1

i run the example “relay_quick_start.py” according to the tutorial “https://docs.tvm.ai/tutorials/relay_quick_start.html#sphx-glr-tutorials-relay-quick-start-py

then there was a warning in the output console as below

WARNING:autotvm:Cannot find config for target=cuda -model=unknown, workload=('dense', (1, 512, 'float32'), (1000, 512, 'float32'), 0, 'float32'). A fallback configuration is used, which may bring great performance regression.

my environment is as below

OS  Ubuntu 1804
GPU GeForce GTX 1060
CUDA 10.0

is there any advice about how to solve it


#2

For some model we upload tuned results to tophub, while for customized model you can always tune by yourself: https://docs.tvm.ai/tutorials/index.html#auto-tuning


#3

can i access the tophub? and what’s the url for the tophub?