Was this trained on the entire pixelprose dataset?
#1
by
xzuyn
- opened
If not, how many of the images were used? Also was any processing done to the captions before training, and could you possibly share the code you used to train?