<h1 id="heading-1">PROJECT_PINAY</h1><h2 id="heading-237">Pinay 1 (1024)</h2><h2 id="heading-237">Pinay 2 (1024)</h2><h2 id="heading-860">Pinay 3 (1024)</h2><h2 id="heading-237">Enjoy, Rate, and share your generations!</h2><h2 id="heading-4016">Intentions</h2>Each LoRa version is a different character with unique and independent datasets. So give em all a try!These LoRas are intended to create Filipinas (Pinay) with very specific physical traits, concentrated almost exclusively on facial features that I find attractive. Thus each LoRa version creates a very unique looking character.<h2 id="heading-861">Info</h2>The only testing I am able to do is on Dezgo so I will include all relevant info associated with that platform. It is as follows:<ul><li>Trigger: Perfect Pinay</li></ul><ul><li>Model: AbsoluteReality 1.8.1</li><li>LoRa Weight: 0.7 to 0.8 For "static" generations</li><li>LoRa Weight: 0.6 to 0.4 for more flexibility in generations</li><li>Resolution: Portrait</li><li>Guidance: 7 to 9</li><li>Sampler: DDIM (this has worked best for me so far, but the others should work as well)</li></ul>I chose to use Dezgo for a few reasons:<ul><li>Free to use</li><li>A lot of options for free use</li><li>No pesky filters that limit 'NSFW' generations </li><li>Frequent updates and additions for free use tools</li><li>Single LoRa support for free use up to 150 MB and SD 1.5</li></ul><h2 id="heading-4018">Technical Stuff</h2><ul><li>Trigger: Perfect Pinay</li></ul>I use the following prompt string in Dezgo to achieve the realism look:<ul><li>8k, raw dslr photo, ultra hyper photo realistic, masterpiece, real life, life like,</li></ul>Some other useful prompts to use are:<ul><li>Small breasts</li><li>Petite</li><li>Slim</li></ul>Example Prompt:<ul><li>8k, raw dslr photo, ultra hyper photo realistic, masterpiece, real life, life like, Perfect Pinay, nude, slim, petite, small breasts,</li></ul><h2 id="heading-10">Tools</h2>I used the free tools provided within the guide listed in the resource section of this page to create all of my LoRas.

Project_Pinay

OUT OF ALPHAThis version has yielded the absolute best results. It's reasonably versatile and super stable. It was trained at 1024 resolution so it really captures the features that were intended.Recommendations:<ul><li>Weight: 0.7 for static results</li><li>Weight: 0.6 to 0.4 for more flexibility.</li></ul>Note: The images shown were generated using Dezgo. I would love to see what you're generating. Please share below and don't forget to rate the LoRa.

Files with SHA256: e2e76a78bb1c166bca67d5ef3c604a7d18adceea745a3bbf425812b5e6c4a874