ModelsLab/Pony: People'S Works V1-V6 - V2 Ani3.1

Base model: Animagine v3.1

batch size=12

dimension=32 conv_dim=8

alpha=16 conv_alpha=2

learning rate: U-net lr=6e-4 Te lr=2e-4

optimizer: AdamW

训练策略：

训练数据集：374张Cvitai上的pony v6图片，393张正则图片

训练总共21个epoch，每个epoch中重复5次训练图片、2次正则图片

使用了 cosine with restarts 调度器，第一个epoch（323步）作为预热；剩余的20个epoch分为5个循环，每个循环中有4个epoch（也就是说每个cycle里训练了20次训练图片）

每张图片训练了5+5×4×5=105次。

Training strategy:

Training dataset: 374 selected pony v6 pics from Civitai, 393 regularization images

Trained 21 epochs, each epoch repeated training training data 5 times and reg data 2 times.

Applied cosine with restarts scheduler, first 1 epoch (323 steps) for warmup and divided rest 20 epochs into 5 cycles, each cycle contained 4 epochs (training images were trained 20 times per cycle).

Trained 5+5×4×5=105 times for one pic totally.