description of input/ouput of the models

[vpenades]

Hi, I've been trying to use the models using another framework, I tried to follow the python code to define de input and output of the two models unsuccessfully.

So far I got:

512 random values > Style model > 512 style values, truncated? > Synthesys model > final image.

Should the random values be between 0 and 1? any additional requirement?

So I need to know the expected values for each input/output, and how to truncate the style values.

[bes-dev]

@vpenades

• Random values are sampling from standard normal distribution N(0, 1).
• We truncate the style values using average style vector
• You can find more information about output values from the code here