Closed gitlabspy closed 4 years ago
Sorry give me some example of your problem. Or you can see my some examples In my each flow-based layer.
https://github.com/MokkeMeguru/TFGENZOO/blob/master/TFGENZOO/flows/inv1x1conv.py#L66-L83
I don't know the case of needing fixed-length batch-size.
Thx! I stop tinkering with the batch_size though.
There's a step of a type of flow in flow steps need to know about the batch_size, how can I pre-build the model so that the
batch_size
will feed into model when instantiating a model class likeglow
in your TFGENZOO_EXAMPLE repo? 😄