Experiment with GAN

Give the python working code in ipynb format for below solution on GAN with all relevant comments and explanations

Task

$1$

: Implement and train a conditional Wesserstein GAN with CIFAR

$1$

$0$

$$dataset$.$ Generate

$2$

$5$

$$new images belonging to 'automobile' class. Report IS and FID scores.

Task

$2$

: Implement and train a SNGAN with CIFAR

$1$

$0$

$$dataset$.$ Generate

$2$

$5$

$$new images. Report IS and FID scores.

Task

$3$

: Implement and train a variant of SAGAN without spectral nornamization and TTUB with CIFAR

$1$

$0$

$$dataset$.$ Generate

$2$

$5$

$$new images. Report IS and FID scores.

Task

$4$

: Implement and train a complete SAGAN with CIFAR

$1$

$0$

$$dataset$.$ Generate

$2$

$5$

$$new images. Report IS and FID scores.

Task

$5$

: Explain your results.