Ani (Grok Companions) - LoRA [NoobAI(Vpred / Eps) , Illustrious ]

🔹 Ani (Grok Companions) - LoRA 🔹

Overview:

Ani is one of xAI’s Grok AI companions — a virtual anime-style girl with a gothic aesthetic. She features long blonde twintails, blue eyes, an off-shoulder black dress, asymmetric thigh-highs, and a playful yet seductive charm. This LoRA captures Ani’s signature look for both her default outfit and underwear.

About LoRA Models:

* For the Vpred model, please download: LoRA_SDXL_noobaiVpredV10_grok_ani_v1_prodigy_d16a8

* For the EPS model, please download: LoRA_SDXL_noobaiEpsV11_grok_ani_v1_prodigy_d16a8

Sample Prompt:

Default Outfits

# 🧚‍♀キャラ1

<lora:LoRA_SDXL_noobaiVpredV10_grok_ani_v1_prodigy_d16a8-000090:1.2>

# 🎨 メタ情報 (Meta)

(ani grok:1.1),

# 👤 キャラクター (Character)

1girl, solo, 

# ✨ 身体的特徴 (Body)

# blunt bangs, 

blonde hair, blue eyes, collarbone, black hair tie, lips, lipstick, long hair, medium breasts, sidelocks, skindentation, thighs, twintails,

# makeup, 

# 🥋 衣装 (Outfit)

alternate costume, goth fashion, gothic lolita, lolita fashion,

# - 上着 (Tops)

black dress, detached collar, dress, frilled dress, lace-trimmed dress, layered dress, off-shoulder dress, pleated dress, puffy short sleeves, short dress, short sleeves, strapless dress,

# - 体幹 (Bodywear)

corset, underbust,

# - 脚衣 (Legwear)

asymmetrical legwear, black thighhighs, fishnet pantyhose, fishnet thighhighs, single thighhigh, skirt, thighhighs, zettai ryouiki,

# - 装飾品 (Accessories)

arm ribbon, belt, belt buckle, black belt, black bow, black choker, black ribbon, bow, choker, dress bow, frills, gloves, jewelry, lace, lace choker, lace trim, lace-trimmed choker, large bow, ribbon,

# 💃 ポーズ・構図 (Pose & Composition)

# arms up, bare shoulders, cowboy shot, feet out of frame, from side, hands up, looking at viewer, looking back, looking to the side,standing, w arms,

# 😊 表情 (Expression)

# blush, closed mouth, expressionless, light smile, smile,

:d, 

# 🖼️ 背景・その他 (Background & Misc)

# black background, gradient background, halloween, purple background, simple background,

white background, 

Underwear Outfits

# 🧚‍♀キャラ1

<lora:LoRA_SDXL_noobaiVpredV10_grok_ani_v1_prodigy_d16a8-000090:1.2>

# 🎨 メタ情報 (Meta)

, (ani grok:1.1),

# 👤 キャラクター (Character)

1girl, alternate costume, solo, 

# ✨ 身体的特徴 (Body Features)

## 💇‍♀️ 頭部 (Head)

blonde hair, black hair tie, long hair, medium hair, sidelocks, twintails,

## 😍 顔 (Face)

blue eyes, 

## 💃 体型 (Body)

bare shoulders, breasts, cleavage, collarbone, medium breasts, navel, skindentation, small breasts, thigh gap, thighs, underboob,

# 👗 衣装 (Outfit)

## 👚 トップス・ドレス (Tops & Dresses)

babydoll, black dress, chemise, dress, lace-trimmed dress, negligee, see-through dress, short dress, 

## 👙 下着 (Underwear)

black bra, black panties, bra, highleg panties, lace bra, lace panties, lace-trimmed bra, lace-trimmed panties, lingerie, panties, string panties, underwear, underwear only,

## 🎀 装飾品 (Accessories)

arm garter, arm strap, armband, asymmetrical legwear, black bow, black choker, black gloves, black thighhighs, bow, bow legwear, bridal garter, choker, detached collar, fishnet thighhighs, fishnets, frilled choker, gloves, lace, lace choker, lace gloves, lace trim, lace-trimmed choker, lace-trimmed legwear, polka dot legwear, see-through, thigh strap, thighhighs,

# 🤸‍♀️ ポーズ・構図 (Pose & Composition)

cowboy shot, looking at viewer, standing,

# 😊 表情 (Expression)

blush, light smile, smile,

# 🖼️ 背景 (Background)

brown background, simple background

TrainingCommand

▼Vpre版
````bash
accelerate launch --num_cpu_threads_per_process 4 sdxl_train_network.py --pretrained_model_name_or_path "C:\ProgramTools\__GenerativeAI\StableDiffusionModels\__SDXL\noobaiXLNAIXL_vPred10Version.safetensors" --train_data_dir "H:\Resources\Resources_AI\__Dataset\forLeaningImages\__Character\Grok_Ani" --output_dir "H:\Resources\Resources_AI\__Models\__Lola\__MyLoRAs\SDXL\Character\grok_ani_vpred" --network_module "networks.lora" --network_dim 16 --network_alpha 8 --network_args "conv_dim=16" "conv_alpha=8" --save_precision "fp16" --xformers --gradient_checkpointing --persistent_data_loader_workers --cache_latents --cache_latents_to_disk --max_data_loader_n_workers 1 --enable_bucket --save_model_as "safetensors" --mixed_precision "fp16" --resolution 1024 --min_bucket_reso 512 --max_bucket_reso 2048 --train_batch_size 3 --max_train_epochs 200 --save_every_n_epochs 10 --optimizer_type "prodigy" --learning_rate 1 --lr_scheduler "cosine" --output_name "LoRA_SDXL_noobaiVpredV10_grok_ani_v1_prodigy_d16a8" --caption_extension ".txt" --max_token_length 225 --keep_tokens_separator="|||" --shuffle_caption --seed 42 --no_half_vae --logging_dir "E:\Creative\Program\Github\__AI\Art\TraningLog\sd_scripts\log" --log_prefix "LoRA_SDXL_noobaiVpredV10_grok_ani_v1_prodigy_d16a8" --fp8_base --optimizer_args "d_coef=0.8" "decouple=True" "weight_decay=0.01"  "use_bias_correction=True" "safeguard_warmup=True" "betas=0.9,0.99" --noise_offset 0.05 --v_parameterization --zero_terminal_snr
````


▼EPS版
````bash
accelerate launch --num_cpu_threads_per_process 4 sdxl_train_network.py --pretrained_model_name_or_path "C:\ProgramTools\__GenerativeAI\StableDiffusionModels\__SDXL\noobaiXLNAIXL_epsilonPred11Version\noobaiXLNAIXL_epsilonPred11Version.safetensors" --train_data_dir "H:\Resources\Resources_AI\__Dataset\forLeaningImages\__Character\Grok_Ani" --output_dir "H:\Resources\Resources_AI\__Models\__Lola\__MyLoRAs\SDXL\Character\grok_ani_eps" --network_module "networks.lora" --network_dim 16 --network_alpha 8 --network_args "conv_dim=16" "conv_alpha=8" --save_precision "fp16" --xformers --gradient_checkpointing --persistent_data_loader_workers --cache_latents --cache_latents_to_disk --max_data_loader_n_workers 1 --enable_bucket --save_model_as "safetensors" --mixed_precision "fp16" --resolution 1024 --min_bucket_reso 512 --max_bucket_reso 2048 --train_batch_size 3 --max_train_epochs 200 --save_every_n_epochs 10 --optimizer_type "prodigy" --learning_rate 1 --lr_scheduler "cosine" --output_name "LoRA_SDXL_noobaiEpsV11_grok_ani_v1_prodigy_d16a8" --caption_extension ".txt" --max_token_length 225 --keep_tokens_separator="|||" --shuffle_caption --seed 42 --no_half_vae --logging_dir "E:\Creative\Program\Github\__AI\Art\TraningLog\sd_scripts\log" --log_prefix "LoRA_SDXL_noobaiEpsV11_grok_ani_v1_prodigy_d16a8" --fp8_base --optimizer_args "d_coef=0.8" "decouple=True" "weight_decay=0.01"  "use_bias_correction=True" "safeguard_warmup=True" "betas=0.9,0.99" --noise_offset 0.05
````