the number of learning iteration or epoch

[ryhhtn]

In the description of paper, the number of trials is 120K.
No other items, epoch or mini batch size, are mentioned

On the other hand, the pretrained model says 1000 epoch.

What criteria can be used to reproduce the same environment as the pre-trained model published in the paper or GitHub Code?

ADGAN/data/keypoint.py

Line 106 in 4dd7064

return 4000

In addition, the paper states that 101,966 pairs were used as train data, but as shown above, the source code is limited to 4,000.
Was this done to limit the number of iterations when learning at 1,000 epoch?