Trained on 335 images of Silver Fox (no repeats) from Kemono Friends on Anything v3.0 Pruned FP32 (however it works on every model)
This time I added the tag "gingitsu" to the training so that it associates "gingitsu" with many features of her, and fixes the association with "silver fox" (especially with its double meaning, however it insists on the animal) and "kemono (in parenthesis!)" messing up the generation
Of course, silver fox (kemono friends) still works
Stats:
335 512x images (no repeats) for 20 epochs
Network dimension: 32
Base model: Anything-V3.0-pruned-fp32.ckpt
Batch size of 3 on RTX 3060
Generation tags: gingitsu, 1girl, solo, animal ears, fox ears, long hair, fox tail, silver hair, black pleated skirt, black gloves, black fur trim, black bowtie, pantyhose, black legwear, long sleeves, black necktie, black undershirt, hair between eyes, very long hair, fur-trimmed sleeves, fox girl, bangs, orange eyes, blue blazer, sidelocks, black loafers, black ears, black tail, white tipped tail
Sample images are 512x512 latent upscaled to 1024x1024 and generated on Anything v3.0 Pruned FP32
描述:
训练词语: gingitsu
名称: gingitsu.pt
大小 (KB): 37116
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: **Detected Pickle imports (3)**
```
torch._utils._rebuild_tensor_v2
torch.HalfStorage
collections.OrderedDict
```
病毒扫描结果: Success
