In order to run a benchmark, one needs to execute the following command:
smlb
. It takes a few parameters. The first one is for telling which model to
run. Supported models can be listed through the help options: smlb --help
.
See the Commands:
section at the end of the output. Each of the models should
have an option for supported frameworks. This can also be checked through the
--help
option. This time however, you need to list the help for the model.
For example: smlb transformer --help
. Look at the Commands:
again.
venekamp / surf-ml_benchmark Goto Github PK
View Code? Open in Web Editor NEWBenchmarking AI/ML