Follow up to my other textual inversion, but this time something that produces safe for work images by default, unless prompted otherwise.
Do note that the trigger words for v1 vs v2 are different, as are the filenames.
The word 'woman' must be used in your prompt somewhere to trigger the best results.
Description
Updated version of the textual embedding. Doesn't necessarily produce 'prettier' pictures, but does bring in more details into the picture with the same settings vis-a-vis the previous version.
Have deliberately put a different trigger word (dgdressed) for this version, in case you want to continue using the previous.