<h1 id="heading-1682"><a target="_blank" rel="ugc" href="http://di.FFUSION.ai">di.FFUSION.ai</a>-tXe-FXAA</h1><h3 id="heading-1683">Trained on "121361" images. <img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/ad54f8db-530b-4d20-97c0-c546c0d7988e/width=525/ad54f8db-530b-4d20-97c0-c546c0d7988e.jpeg" /> </h3>Enhance your model's quality and sharpness using your own pre-trained Unet. The text encoder (without UNET) is wrapped in LyCORIS. Optimizer: torch.optim.adamw.AdamW(weight_decay=0.01, betas=(0.9, 0.99))Network dimension/rank: 768.0 Alpha: 768.0 Module: lycoris.kohya {'conv_dim': '256', 'conv_alpha': '256', 'algo': 'loha'}Large size due to Lyco CONV 256<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/58fb1eb8-2eaf-416d-9511-80c996f9ecc1/width=525/58fb1eb8-2eaf-416d-9511-80c996f9ecc1.jpeg" /><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/2858847c-3f55-48c5-b127-472716acaafc/width=525/2858847c-3f55-48c5-b127-472716acaafc.jpeg" />For a1111 Install <a target="_blank" rel="ugc" href="https://github.com/KohakuBlueleaf/a1111-sd-webui-lycoris">https://github.com/KohakuBlueleaf/a1111-sd-webui-lycoris</a>Download <a target="_blank" rel="ugc" href="http://di.FFUSION.ai">di.FFUSION.ai</a>-tXe-FXAA to /models/LycorisOption1:Insert &lt;lyco:<a target="_blank" rel="ugc" href="http://di.FFUSION.ai">di.FFUSION.ai</a>-tXe-FXAA:1.0&gt; to prompt No need to split Unet and Text Enc as its only TX encoder there.You can go up to 2x weightsOption2: If you need it always ON (ex run a batch from txt file) then you can go to settings / Quicksettings listadd sd_lyco<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/cdf94aee-f8c9-4d9a-a3e9-ed2e6cdbb955/width=525/cdf94aee-f8c9-4d9a-a3e9-ed2e6cdbb955.jpeg" />restart and you should have a drop-down now 🤟 🥃 <img src="https://user-images.githubusercontent.com/50985923/238659318-db7ea36c-4d59-4c9f-83c0-e0cc09741e64.png" alt="image" />More info: "ss_text_encoder_lr": "1e-07", "ss_keep_tokens": "3", "ss_network_args": { "conv_dim": "256", "conv_alpha": "256", "algo": "loha" }, "img_count": 121361 } "ss_total_batch_size": "100", "ss_network_dim": "768", "ss_max_bucket_reso": "1024", "ss_network_alpha": "768.0", "ss_steps": "2444", "sshs_legacy_hash": "539b2745", "ss_batch_size_per_device": "20", "ss_max_train_steps": "2444", "ss_network_module": "lycoris.kohya", This is a heavy experimental version we used to test even with sloppy captions (quick WD tags and terrible clip), yet the results were satisfying.Note: This is not the text encoder used in the official FFUSION AI model.

Files with SHA256: a6b0eeb89cef27a927b6acd5ee035c5ef047abaf2801e4396587c61e02814c01