Closed traines3812 closed 1 year ago
well, when generating passwords you have 4 options:
A pretrained PassGAN model trained on the RockYou dataset : directory pretrained 10 chars A pretrained PassGAN model trained on the darkc0de+openwall+xato-net-10M dataset 11 chars : directory commonu11 A pretrained PassGAN model trained on the crackstation-only-human dataset 10 chars : directory csho10
The 4th option is to train your own dictionary possibly with length 10-11 chars max each password with custom directory
When generating passwords with sample.py you need to choose one of this datasets and adjust all parameters according to the dataset of your choice. Most of times pretrained datasets are good enough to generate very big dictionaries.
You can read more info in readme file and on the article : (https://www.tuxrincon.com/es/blog/passgan-como-generar-archivos-de-passwords-con-samplepy/)
thank you
Hello beta6 thank you again for help on original issue. THis one is more of a q/a.
I am currently training PassGAN with some custom wordlists using the example that has been around since the first iteration of passgan
python train.py --output-dir output --training-data data/train.txt
How would I go about making the custom wordlists?
Would I follow one of these examples?
python sample.py \ --input-dir csho10 \ --checkpoint csho10/checkpoints/checkpoint_150000.ckpt \ --output gen_passwords.txt \ --batch-size 1024 \ --num-samples 100000000
but change the input dir to output to match that of the custom wordlists and just choose a checkpoint?