An experimental LoRa, trained on <a target="_blank" rel="ugc" href="http://pixai.art">pixai.art</a>. Works best with realistic models.Trigger word is funky_cars.Chose a low strength, i.e. 0.5 and you are able to have a little influence on what car could be created .Increase the strength up to 1.0 and you'll get any (very, very ill) car, that could be manufactured between 1950 and 1960.Don't expect too much from the quality, training a LoRa on <a target="_blank" rel="ugc" href="http://pixai.art">pixai.art</a> might sometimes lead to proper results, but unless there's nothing much to adjust for training, it's not really predictable, what you'll get.I've been tinkering round with many Checkpoint models.kawaiiRealistic_v01realisticVisionV60B1_v51HyperVAEicbinpICantBelieveIts_v8Those ckpt models generate good results with a LoRa strength around 0.5 - 0.6.However, do whatever you want with this LoRa and have fun fiddling around with it.

Files with SHA256: dfc50badd6aa40e52df4b04551957cfc5271d4c8be7cd674ea8de9f3814d7652