For best outputs:

Version 1.8:

Sampler: DPM++ 2S a Karras.

Steps: 35, cfg 7, CLIP skip 1.

VAE: blessed2_fp16.safetensors

Pos prompt: add this to the end of your prompt for more realism "realistic, real photo, realistic skin texture, skin imperfections, skin pores, sharp details, bright light, 8k, HDR, f.1/2, 70mm lens, Kodak, Analog style OverallDetail, epiC35mm, , ", being the LoRa's "epiC35mm" and "add_detail" included in my v1.8 example runs.

Neg prompt: "nsfw, nipples, nudity, portrait, focus on face, text, logo, (worst quality:2), (low quality:2), (normal quality:1.5), lowres, ((monochrome)), ((grayscale)), sketch, ugly, morbid, deformed, logo, text, bad anatomy, bad proportions, unrealistic, low resolution, big eyes, epiCNegative, realisticvision-negative-embedding", being the embeddings "epiCNegative" and "realisticvision-negative-embedding" included in my v1.8 example runs.

Version 1.5:

Sampler: DPM++ 2M Karras.

Steps: 30, cfg 7, CLIP skip 2.

VAE: none required.

Pos prompt: add the LoRa "add_detail" with a weight of 0.8 at the end of your prompt (included in both versions' example runs).

Neg prompt: "(bad-hands-5:1.5), realisticvision-negative-embedding, ((nude:2)), ((nipples:2)), ((NSFW:2)), ((topless:2))", being the embeddings "bad-hands-5" and "realisticvision-negative-embedding" included in my v1.5 example runs.

What changes from v1.5 to v1.8: input_image feature, overall quality and realism increased.

Checkpoint Merge.