PonyモデルならばLoRAを作るまでもなくプロンプトで作成できるので作るつもりはなかったが画像の前処理用<a rel="ugc" href="https://github.com/NEXTAltair/Lorayougazouwoiroirosurusukuriputo">スクリプト</a>のテストのために作成構図が固定されすぎていまいち存在意義が分かりづらい"I wasn't planning on creating a LoRA because the Pony model can be created with prompts without needing one. However, I created one for testing the image preprocessing script. The composition is too fixed, making it difficult to understand its significance."

ass cutout backless panties - Clothing XLPony

<ul><li>caption_dropout_rate: 0.05</li><li>enable_bucket: true</li><li>epoch: 50</li><li>fp8_base: true</li><li>gradient_accumulation_steps: 1</li><li>gradient_checkpointing: true</li><li>learning_rate: 1.0</li><li>lr_scheduler: "cosine"</li><li>max_train_steps: 7350</li><li>optimizer_type: "Prodigy"</li><li> optimizer_args = [ "decouple=True", "weight_decay=0.01", "d_coef=0.8", "use_bias_correction=True", "safeguard_warmup=False", "betas=0.9,0.99",]</li><li>resolution: "1024,1024"</li></ul>

Files with SHA256: f22dfbc16eef8f0df5f721d739506a32dc980aa14c9512b722826be52f162032