V2 uses the same keyword 'AC1DG' as v1 but has also been captioned and tends to produce better but slightly less chaotic (hard to explain) images. The test images shown were generated with the exact same parameters as the V1 model so you can compare.
V2 uses the same keyword 'AC1DG' as v1 but has also been captioned and tends to produce better but slightly less chaotic (hard to explain) images. The test images shown were generated with the exact same parameters as the V1 model so you can compare.