1. Setup and Initialization
The process began on September 8, 2025, at 06:01:09 AM, when the training job was submitted. The system, running on an NVIDIA CUDA device, quickly created a dedicated workspace for the job.
The training was configured using the kohya engine, a popular tool for fine-tuning Stable Diffusion models. The system loaded key settings from a training_config.toml file, including:
Learning Rate: A
unetLRof0.0005and atextEncoderLRof0.00005, which controls how quickly the model's parameters are updated.Optimizer:
Adafactorwas chosen as the optimization algorithm.Model Type: The
networks.loramodule was imported, confirming that a Low-Rank Adaptation (Lora) model was being trained.Dataset: The dataset consisted of 34 images, each with a corresponding
.txtcaption file. The captions were formatted as simple tags.
2. Training Execution
The training loop began, with the system calculating its total steps and batches.
Batches and Epochs: The training used a
trainBatchSizeof 4 images per step. With 34 images and anum_repeatsof 58, the total number of training images was34 * 58 = 1972. This resulted in495batches per epoch and a total of9900optimization steps for 20 epochs.Data Augmentation: To prevent the model from overfitting to the small dataset, data augmentation techniques were used.
flip_aug(horizontal flipping of images) andshuffle_caption(randomizing the order of tags) were enabled.Monitoring Progress: The
stdoutlogs showed the training progressing epoch by epoch, saving asafetensorscheckpoint file after each one. For example,saving checkpoint: ...000001.safetensorswas logged after the first epoch.
3. Checkpointing and Finalization
As the training progressed, the system continuously saved its work.
Intermediate Models: The
workerlogs show that the system was uploading intermediate models to the orchestrator. For example, at one point, it logged that it had "uploaded 7 models and 21 sample images so far."Sample Image Generation: The logs also show the model generating sample images at various points using the prompts provided in the
sample-images.jsonfile. The prompts included phrases like "my father" and "my father and his friend just1n, kissing."
The process concluded on September 8, 2025, at 02:31:17 PM, when the status was updated to Ready, indicating that the model was fully trained and the final safetensors file was available for use.
Description
Subject and Identity
Characters:
a well built obese bodybuilder,a hairy thick pectorals dad,old father rugged dad,a muscular middle-aged man,hairy old man,dad body,dilf,middle-aged man,mature male,asian man,everyday humanAttributes:
1boy,1man,solo male focus,male focus,bara,middle-aged,plump,fat man,manly
Physical Appearance
Body Type:
muscular,muscular male,beefy dadbod,hairy chested,strong upper body,thick,plump,fat,small belly,slightly overweight,hairy body,hairy arms,hairy legs,abs,six-pack abs,stomachBody Parts:
thick pectorals,large pectorals,hairy body,hairy arms,hairy legs,abs,stomach,navel,nipples,puffy nipples,armpit crease,hairy pubes,hairy small testicles,heavy balls,veins in cock,veiny arms,penis,balls,cock glans,beard,stubble,mustacheFacial Features:
short hair,black hair,grey hair,thick eyebrows,open mouth,closed mouth,teeth,parted lips,lips,nose,brown eyes,green eyes,strong jawline,clean-shaven face,smooth skinModifications:
tattoos on shoulder,nipple rings,pierced nipples,prince albert,PA piercing,PA chain
Expressions and Actions
looking at viewer,smirk,smile,grin,blush,neutral expression,confident expression,clenched teeth,squatting,standing,sitting,holding his cock with his hand,hand on hip,casual and confident pose,relaxed atmosphere,cheerful expression
Clothing and Accessories
shorts,pants,black pants,white pants,grey shorts,gray pants,tuxedo,black tuxedo,suit,black suit,grey jacket,open clothes,open jacket,shirtless,topless male,white tank top,sleeveless shirt,bare shoulders,bare arms,ring,watch,smartwatch,wedding ring,black-framed eyewear,rimless eyewear,sneakers,socks,black backpack,pink short-sleeve shirt,pink polo shirt,floral pattern shirt
Explicit Content (NSFW)
nsfw,explicit,cock,horsecock,equinecock,thick flared equine-penis horsecock,erect cock,circumcised,uncircumcised,cum on the tip of his cock,urethral fingering,urethral insertion,urethral bulge
Setting and Environment
Location:
indoors,outdoors,in a room,bedroom,bathroom,church,street,city,carnival,funfair,amusement park,living room,urban environment,urban streetTime and Weather:
daytime,night,night time,night sky,raining,overcast,wet streetObjects:
bed,sofa,curtains,window,desk,car,motor vehicle,ferris wheel,yellow crane,street lamp,street sign,towels,sink,mirror,posters,wooden furniture,framed pictures,potted plant,bicycle
Photography Style and Details
Style:
realistic,realistic photograph,candid,natural,unfiltered,casual,amateur shot,street photographyTechnical:
soft lighting,natural lighting,shallow depth of field,blurry background,rule of thirds,eye-level,medium shot,medium close-up,Canon EOS 5D Mark IV,Canon EOS 5D Mark III,Canon EOS Rebel T6,smartphone camera,no watermark,no filters,no jpeg artifacts,f/1.8,f/2.8,ISO 400,ISO 800,ISO 3200,1/60 sec,1/200s,1/250s