This is the SD 1.5 version of Pony V6 fine-tuned at native 1024px on ~5000 hand-selected images (some of them borrowed from my Zootvision model's datasets). Each image was captioned both with Florence-2 Large "More Detailed Mode" rich captions and also Booru tags from WD-VIT-V3. Use this the same way you'd use Pony V6 SD 1.5 normally, and just uh, enjoy the pretty objectively better overall aesthetics.Important notes:<ul><li>in A111, use Clip Skip 1 (NOT 2) and in Comfy just do not use the "Clip Set Last Layer" at all, with this</li><li>DO NOT use any VAE besides the one that is baked into the checkpoint, it will fry the image</li><li>You will probably get worse results from weird "in between" resolutions like 720x1280 than you will from standard ones like 768x1024 and 832x1216</li></ul>Basic positive prompt: <code>score_9, source_whatever, rating_whatever, your tags or natural language description here</code>.Basic negative prompt: <code>score_3_up, score_4_up, score_5_up, sketch, (simple background:1.2)</code>.Recommended steps / sampler CFG: typically Euler Ancestral at CFG 7.0 with around 25 - 35 steps is a good starting place. The DPM++ SDE GPU family of samplers can also be good with this at lower CFG (4.0 - 5.0) if you're going for realism in particular.Generating at 512x512 is NOT recommended, as this model was originally trained by AstraliteHeart at 768px, and my additional training was entirely done at 1024px.

Better Pony Diffusion V6 For SD 1.5

Flux dataset integrated as well as a smaller one for a few other miscellaneous things. Best version ever for sure IMO.

betterPonyDiffusionV6_v6.safetensors

v5.75

v5.5

v5.0

v4.0

v3.0

v2.0

v1.0

This is the SD 1.5 version of Pony V6 fine-tuned at native 1024px on ~5000 hand-selected images (some of them borrowed from my Zootvision model's datasets). Each image was captioned both with Florence-2 Large "More Detailed Mode" rich captions and also Booru tags from WD-VIT-V3. Use this the same way you'd use Pony V6 SD 1.5 normally, and just uh, enjoy the pretty objectively better overall aesthetics.Important notes:<ul><li>in A111, use Clip Skip 1 (NOT 2) and in Comfy just do not use the "Clip Set Last Layer" at all, with this</li><li>DO NOT use any VAE besides the one that is baked into the checkpoint, it will fry the image</li><li>You will probably get worse results from weird "in between" resolutions like 720x1280 than you will from standard ones like 768x1024 and 832x1216</li></ul>Basic positive prompt: score_9, source_whatever, rating_whatever, your tags or natural language description here.Basic negative prompt: score_3_up, score_4_up, score_5_up, sketch, (simple background:1.2).Recommended steps / sampler CFG: typically Euler Ancestral at CFG 7.0 with around 25 - 35 steps is a good starting place. The DPM++ SDE GPU family of samplers can also be good with this at lower CFG (4.0 - 5.0) if you're going for realism in particular.Generating at 512x512 is NOT recommended, as this model was originally trained by AstraliteHeart at 768px, and my additional training was entirely done at 1024px.

Flux dataset integrated as well as a smaller one for a few other miscellaneous things. Best version ever for sure IMO.

Files with SHA256: ef312250f0e2e70e9138e1b61410c6c175001e19e2472e4400a0a1f63fcb3800