Improving inference time

[joshwa71]

We are looking to run this model and reduce the processing time. Is there any way to tune hyperparameters and/or parallelise inference so that we can leverage more compute?

[Xzzit]

Adding the --no_flip option to the command line in my case results in 5s improvement in the inference phase. Although there is a minor reduction in image quality, the time savings make it a worthwhile trade-off.