<h3 id="training-data-5mblkhxod">Training Data</h3>concepts: baseskin, blackchroma, bluechroma, greenchroma, pinkchroma, redchroma, tealchroma, whitechroma, yellowchromatags: armlet, bare shoulders, breasts, cleavage, detached collar, elbow gloves, eyes, facing away, frills, from behind, from side, full body, holding, long hair, looking at viewer, mermaid, mermaid tail, pumpkin hat ornament, solo, staff, transparent background, witchhat, witchdressTrained on 45 images @ 45 epochs (2025 steps total) of in-game model 5 of each skin variant baseskin, blackchroma, bluechroma, greenchroma, pinkchroma, redchroma, tealchroma, whitechroma, yellowchromaTraining Data here: <a target="_blank" rel="ugc" href="https://imgur.com/a/jcZRm1U">https://imgur.com/a/jBOVxDy</a>Weight Data Type: bfloat16Override Unet Data Type: bfloat16Override Text Encoder 1 Data Type: bfloat 16Output Data Type: bfloat 16Latent Caching: TrueClear Cache Before Training: TrueOptimizer: AdamWBeta1: 0.9EPS: 1e-30Stochastic Rounding: TrueBeta2: 0.999Weight Decay: 0.01Fused Back Pass: TrueLearning Rate Scheduler: ConstantLearning Rate: 0.0003Learning Rate Warmup: 200Learning Rate Cycles: 1Epochs: 45Batch Size: 1Accumulation Steps: 1Learning Rate Scheduler: NoneAttention: XFORMERSEMA: OFFGradient checkpointing: TrueTrain Datatype: bfloat16Fallback Train Datatype: bfloat16Autocast cache: trueResolution: 1024Train Text Encoder 1: TrueStop Training After 45 EpochsText Encoder 1 Learning Rate: 3e-06Text Encoder Clip Skip: 2Train UNet: TrueStop Training After 45 EpochsUnet Learning Rate: 1e-05LoRA Rank: 16LoRA Alpha: 1.0Dropout probability: 0.0LoRA Weight Datatype: float32The LoRA seems a little off x.xThe concept tags don't seem to properly work for the coloring &amp; dress.I might work on the tags later on... the hair &amp; face comes out fine but the dress, gloves, hat, &amp; staff doesn't most of the time for some reason...I think the issue is the chromas were tagged the same witchdress &amp; witchhat when I should've tagged them appropriately...The pumpkin hat ornament tag has issues with placing multiple pumpkins in the image outside of the hat area.Everything should be adjustable hair, breast size, dress, eye color, gloves, &amp; hat as they were tagged.I suggest removing the tag mermaidtail if you want to generate only halfbusts.It's my second lora I've trained @_@ so if anyone has any tips I'd like them ❤️

Bewitching Nami [Pony v6]

Bewitching Nami skin from League of Legends

Files with SHA256: 0fd973c71b420f3cec6d63f49d2a7f33e016dbbee08a466c12cb7cd41266586b