Revert unet to cb6f860 71400 steps, after fixing min image tracking

v1 model retrained on new master dataset to 4800 steps

another 18800 steps on 118400 ckpt, starting to deform some subjects. approx 55 epochs of the early dataset.

reset 118400 branch to 0 steps and then did 11800 steps on a photo dataset for realism finetuning. hands got harmed.

71400 steps, after fixing min image tracking

57800 steps

prev checkpoint at 42600 steps carried forward to 52000 steps without offset noise, improved coherence and reduced crust further

36000 + 21800 + 14200 + 42600 steps, 118400 total - continued training from 30k steps on to 42600 steps using cropped squares with conditional inputs

86600 steps total, 3000 more than last with focus on timesteps 0-400

36000 + 21800 + 14200 + 11600 = 83600 steps total, post-v1 training covering timesteps 400-1000

36000 + 21800 + 14200 = 72600 steps total, released as v1

36000 + 21800 steps MJ 5.2 (a little more than one epoch) = 57800 steps with min-SNR gamma

36000 steps, mj 5.2 general

28250 steps of min-SNR gamma training on MJ 5.2, LAION-COCO, BG20K, LAION-HD (EXIF), some LAION-Aesthetics

checkpoint-17400, one-shot trained