For best outputs:

Sampler: DPM++ 2M Karras.

Steps: 30, cfg 9, CLIP skip 2.

VAE: kl-f8-anime2_fp16.safetensors.

Pos prompt: add the LoRa "more_details" with a weight of 0.7 at the end of your pos (found in my example runs).

Neg prompt: "(By bad artist -neg:0.9), ((duplicated)), nsfw, watermark, text, error, blurry, jpeg artifacts, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, artist name, (worst quality, low quality:1.4), bad anatomy", being the embedding "By bad artist -neg" found in my example runs.

Checkpoint Trained.