
In my never ending quest to cast my own movie here is Orlando Bloom. All photos are from movies (mostly LotR) not press events so they are grittier and more dramatic in the outputs.
Works very well in my testing and plays well with other embeddings. All raw text2img examples.
Enjoy and please comment pics of what you make with it!
描述:
Version 1. Trained on 25 images with a batch size of 10 and gradient accumulation of 3. This version leans pretty heavily on the long blonde straight hair, so in a future version I will include more training data with hair variations.
Learning rate: 0.05:10, 0.02:20, 0.01:60, 0.005:200, 0.002:500, 0.001:3000, 0.0005
训练词语: orlandobloom
名称: orlandobloom.pt
大小 (KB): 25
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success