Eimi Isami from Phase Connect.YT: <a target="_blank" rel="ugc" href="https://www.youtube.com/@eimiisami">https://www.youtube.com/@eimiisami</a>X: <a target="_blank" rel="ugc" href="https://twitter.com/eimiisami">https://x.com/eimiisami</a><ul><li>Flux Dev &amp; Schnell - Hamster gets the flux treatment. Weights 0.75 - 1 recommended.</li><li>SDXL v2.0 - Corrected a huge typo in the dataset and added more training images.<ul><li>For the eye color, using "faded fushia gradient eyes" seemed to work the best.</li></ul></li><li>Added a SDXL v1.0 version trained using <a target="_blank" rel="ugc" href="https://civarchive.com/models/199934/flatpiececorexl?modelVersionId=325300">FlatpieceCoreXL</a> + <a target="_blank" rel="ugc" href="https://civarchive.com/models/199948?modelVersionId=224985">FlatpieceXL VAE</a><ul><li>Upper body focused. Eye color isn't 100% accurate.</li></ul></li></ul><ul><li>V2.0 (SD1) - Trained on multiple images using model <a target="_blank" rel="ugc" href="https://civarchive.com/models/23900/anylora-checkpoint">anylora</a> + <a target="_blank" rel="ugc" href="https://civarchive.com/models/118561/anything-kl-f8-anime2-vae-ft-mse-840000-ema-pruned-blessed-clearvae-fp16cleaned">anything kl-f8-anime2 vae</a></li><li>V1.0 (SD1) Lora was trained on a single image of Eimi but its apparently good enough to post here.</li></ul>

Eimi Isami - Phase Connect (XL/Flux)

Fixed a massive typo in the dataset + improvements using what I've learned recently

EIMI-XLv2-t1-000002.safetensors

SDXL v2.0

Flux Schnell

Flux Dev

SDXL V1.0

v2.0

v1.0

THIS IS A REPRINT!Go to original author to check their other models or to support them. <hr /> Eimi Isami from Phase Connect.YT: <a target="_blank" rel="ugc" href="https://www.youtube.com/@eimiisami">https://www.youtube.com/@eimiisami</a>X: <a target="_blank" rel="ugc" href="https://twitter.com/eimiisami">https://x.com/eimiisami</a><ul><li>SDXL v2.0 - Corrected a huge typo in the dataset and added more training images.<ul><li>For the eye color, using "faded fushia gradient eyes" seemed to work the best.</li></ul></li><li>Added a SDXL v1.0 version trained using <a target="_blank" rel="ugc" href="https://civitai.com/models/199934/flatpiececorexl?modelVersionId=325300">FlatpieceCoreXL</a> + <a target="_blank" rel="ugc" href="https://civitai.com/models/199948?modelVersionId=224985">FlatpieceXL VAE</a><ul><li>Upper body focused. Eye color isn't 100% accurate.</li></ul></li></ul><ul><li>V2.0 (SD1) - Trained on multiple images using model <a target="_blank" rel="ugc" href="https://civitai.com/models/23900/anylora-checkpoint">anylora</a> + <a target="_blank" rel="ugc" href="https://civitai.com/models/118561/anything-kl-f8-anime2-vae-ft-mse-840000-ema-pruned-blessed-clearvae-fp16cleaned">anything kl-f8-anime2 vae</a></li><li>V1.0 (SD1) Lora was trained on a single image of Eimi but its apparently good enough to post here.</li></ul>

Files with SHA256: 84bbc887834cf2521721ed2cda1d92c9f70d4fb4e5b4f6d0ef0862982b5837d8