PixAI LoRA Training Settings: Advanced Training Guide (2026)

Summary Flow Images Multiple characters Fixed settings Basic Settings Descriptions Parameters kohya mapping Credits & time Checklist Publishing FAQ

Guide · Advanced Training

PixAI LoRA Advanced Training Guide

Steps in the flow

3

Default Training Steps

325

Images

10–100

Training time

2–6 h

Advanced Training gives you direct control over how PixAI trains your LoRA. You can review and edit the description for every image, and set Training Steps, learning rate, detail capacity (Rank), Alpha and batch size yourself. Every default matches Basic Training, so you only change what you want to test.

PixAI LoRA training entry screen with the Advanced Training card highlighted
Choose Advanced Training on the LoRA training screen

Keep Advanced Training open in another tab and follow along with this guide.

Already training LoRAs with kohya or sd-scripts? Three things work differently here

StepsTraining length is set as total steps. There are no epochs or repeats.

AlphaAlpha can only be half, equal to, or double your Rank.

CaptionPixAI writes the image descriptions first. You edit them afterwards.

Quick summary

Steps

Training length is total steps

Training Steps range from 50 to 800, default 325. No epochs or repeats.

Learning rate

One learning rate

Default 6e-4, applied to the whole network. You can't set it per module.

Alpha

Three Alpha options

Half, equal to, or double your Rank, with half as the default. Credit cost depends mainly on Training Steps and batch size. Learning rate, Rank and Alpha don't change it.

Caption

Descriptions are generated first

Written in natural language, not Danbooru tags. You can edit each one afterwards.

Trigger words

Added to every description automatically

Enter them once in Basic Settings and PixAI puts them at the start of each image's description.

Credits

Descriptions and training are billed separately

Described images aren't charged again, a failed training isn't charged, and deleting described images doesn't refund their description cost.

How the training flow works

Advanced Training has three steps: Basic Settings, Descriptions and Advanced Parameters. The order is close to most community LoRA workflows. The main difference is that PixAI writes the captions, and some settings lock once the training is created.

1

Basic Settings →

LoRA name, type, base model and trigger words

2

Descriptions →

Add images, generate descriptions, then edit them one by one

3

Advanced Parameters →

Set steps, learning rate, Rank, Alpha and batch size, check the cost, start training

Descriptions save automatically. If you leave the page, pick up again from Training tasks. You can close the page while descriptions are generating, while the training is queued, and while it runs.

Basic Settings

PixAI Advanced Training Basic Settings screen with LoRA name, trigger words, LoRA type and base model
Enter the LoRA name and trigger words, then choose the LoRA type and base model

Descriptions

PixAI Advanced Training Descriptions step with the image upload area and description tips
Add images from your device, generation history or a past training
PixAI Advanced Training image set with images not described yet and the Describe Automatically button with its credit cost
Click Describe Automatically. The cost appears next to the button
PixAI Advanced Training Edit description dialog with trigger words added automatically
Review and edit each description. Trigger words are added for you

Advanced Parameters

PixAI Advanced Training Advanced Parameters screen with Training Steps, learning rate, Rank, Alpha and batch size
Set Training Steps, learning rate, Rank, Alpha and batch size, check the credits, then start training

Preparing your images

Your image set decides what the LoRA learns. Keep the features you want the LoRA to remember consistent in every image, and let everything else vary. That contrast is how the model tells the LoRA itself apart from things that just happened to be in the picture. For a character LoRA:

GoodBadThe stick figures in the bad examples are illustrations

Angles

Include front, side and back views, plus high and low angles. A set with only front views tends to fall apart when you generate the character from the side or back.

PixAI LoRA dataset example, angles: front, side and back views compared with front views only

Poses and actions

Mix standing, sitting, walking, reaching and other poses. If every image shows the same pose, the LoRA tends to lock onto it.

PixAI LoRA dataset example, poses: standing, sitting, walking and reaching compared with four standing images

Framing and distance

Combine close-ups, half-body and full-body shots. With close-ups only, the model has a hard time learning the outfit and body shape.

PixAI LoRA dataset example, framing: close-up, half-body and full-body compared with close-ups only

Expressions

If you want to control expressions when you generate, include images with different expressions.

PixAI LoRA dataset example, expressions: smiling, sad and surprised compared with neutral faces only

Backgrounds

Vary the backgrounds as much as you can. If most images share one background, the LoRA learns the background too.

PixAI LoRA dataset example, backgrounds: park, room, sky and street compared with the same room every time

Image quality

Pick clear images with no watermarks or text. Crop multi-panel images and comic pages into single images first.

PixAI LoRA dataset example, image quality: clean single images compared with watermarks, text, multi-panel and blur

No near-duplicates

Leave out images that are almost identical, such as the same picture at different sizes. Duplicates get overlearned.

PixAI LoRA dataset example, duplicates: three different images compared with one image at three sizes

Character design

Use one version of the character design. Remove watermarks, signatures and speech bubbles where you can, drop blurry or broken images, and crop away large empty areas.

PixAI LoRA dataset example, character design: one consistent design compared with an old design, a speech bubble and empty space

Style LoRAs work the other way round: vary the characters, scenes and subjects as much as possible, and keep only the art style consistent.

Training several characters in one LoRA

To teach one LoRA two characters, say Character A and Character B, your image set needs three kinds of images: A alone, B alone, and A and B together.

What the image set needs

PixAI multi-character LoRA dataset example: Character A alone, Character B alone, and both together

10+ images per group Groups about the same size

With solo images only, the model struggles to draw both characters in one scene and often blends their features. With group images only, it's hard to generate either character alone. The groups don't need identical counts, but each needs enough images: at least 10 per group, with the groups roughly the same size.

Each LoRA has one trigger word field, so you tell the characters apart by how you name them in the descriptions. Pick a fixed name for each character and use it in every image. Leave out features that never change (hair color, eye color) so they attach to each name. In group images, say where each character is and what each one is doing. When you generate, write the name in your prompt to choose who appears. For the generation side, see our multi-character LoRA generation guide.

Settings PixAI handles for you

These are fixed and can't be changed.

Training resolution

The base model sets the training resolution. Each image needs at least 512×512 pixels in total (about 262,000 pixels) and an aspect ratio between 1:3 and 3:1. Images outside these limits can't be added.

Because the check uses total pixels, an image can pass even if one side is shorter than 512. A 400×700 image has 280,000 pixels, so it's accepted. A 1200×300 image has enough pixels, but its 4:1 ratio is outside the limit, so it's rejected.

PixAI LoRA training resolution rules: 512×512 baseline, 400×700 accepted, 1200×300 rejected
Total pixels and aspect ratio decide whether an image is accepted

Image sources and formats

Upload from your device, pick from your generation history, or import from a past training. PNG, JPG and WebP are supported. Each training needs 10 to 100 images.

Network type

Standard LoRA only. LyCORIS variants such as LoCon and LoHa aren't available.

STEP 1

Basic Settings

Enter a LoRA name, choose the LoRA type (Character, Outfit, Art style or Something else), pick the base model (Tsubaki.2 or Tsubaki.3), and set your trigger words.

You can come back to this step any time before you submit. After you submit, you can still cancel while the training is queued: it turns back into a draft and you can edit this step again. Once training starts, the LoRA type, base model and trigger words are locked, and you'll need a new training to change them. You can rename the LoRA after publishing from Edit Model. Trigger words are added to every description during training, so double-check them before you start.

Once training starts

  • LoRA name Editable after publishing
  • LoRA type Locked
  • Base model Locked
  • Trigger words Locked

Trigger words

Enter trigger words once in Basic Settings. PixAI adds them to the start of every description during training, so you don't need to put them in folder names or caption files as with kohya. To use the LoRA, include the trigger words in your prompt. They can't be changed after you submit. Only letters a–z, numbers, spaces and , . _ - are allowed. There's no minimum length, the limit is 256 characters, and consecutive spaces or spaces at the start or end aren't allowed. For tips, read our LoRA trigger words guide.

PixAI Advanced Training trigger words field in Basic Settings
The trigger words field in Basic Settings

STEP 2

Descriptions

Upload images from your device, pick them from your generation history, or import them from a past Advanced Training. Imports bring their descriptions along, so there's nothing to regenerate and no description cost. You can still edit them after importing. Datasets from Basic Training can't be imported into Advanced Training. For more on reusing image sets, see our multi-version LoRA guide.

  • You can import a dataset as long as its Advanced Training draft still exists. If you delete a LoRA made with Advanced Training, its images and descriptions stay available for import.
  • Deleting the draft is different: the dataset goes with it, and you can no longer import from it. If you may want to retrain with the same images later, keep the draft or back up the images. Otherwise the descriptions have to be generated again and charged per image.

Image descriptions (captions)

Descriptions are the equivalent of kohya caption files. PixAI's description model writes a natural-language description for every image instead of Danbooru tags. Each description has to be generated before you can edit it. Your edits are saved separately, and you can restore the generated version at any time.

A description tells the model what's in each image. During training, the trigger words go in front of every description. Anything a description mentions is treated as something you can change with your prompt. Anything it never mentions gets absorbed into the trigger words and becomes part of the LoRA. So leave out the features you want the LoRA to keep, and describe the things you want to swap when you generate.

PixAI Advanced Training Edit description dialog showing a generated natural-language description with trigger words added automatically
The Edit description dialog after automatic descriptions: plain-language text you can edit and save

What a description does (illustration)

trigger words A girl sitting on a park bench, wearing a red coat and smiling.

Dark: trigger words added by PixAI · Blue: things you wrote, which you can swap in your prompt · Not written (hair color, eye color, signature accessories) → absorbed into the trigger words and fixed in the LoRA

The tips shown in the interface:

  1. Describe only what you can see. Skip personality, backstory and anything off screen.
  2. Describe things you may want to change later, such as outfits, and include images with different outfits.
  3. When more than one person is in the image, make it clear who is doing what. Avoid phrases like "one of them".
  4. Leave out features that never change, like hair color, eye color and signature accessories, so they become part of the character.
  5. Use the same word for the same thing across the whole image set. If you pick "sitting", keep using "sitting".
  6. Don't type the trigger words yourself. They are added to the front of every description automatically.

Every image needs a description before you can move on. If any image is still not described, the interface asks you to finish first.

STEP 3

Advanced Parameters

Set Training Steps, learning rate, detail capacity, Alpha and update smoothing, check the cost, then click Start training. While the training is queued you can still cancel it from Training tasks: the training cost is refunded in full, but description costs aren't. The task becomes a draft again, and you can edit the images, descriptions and Basic Settings before resubmitting. Once training has started, it can't be cancelled.

Parameter by parameter

Training Steps (training length)

Affects credits

max_train_steps

325 Default · 50–800 in steps of 5

PixAI Advanced Training length setting with the Training Steps slider at the default of 325, range 50 to 800
Training length in the interface, default 325 steps

Training Steps set how many times the model studies your images. The default is 325, adjustable from 50 to 800 in steps of 5. The interface suggests leaving it at 325 unless your image set is unusually large or small.

There are no epochs or repeats, and every image has the same chance of being picked. If you think in kohya terms, estimate how often each image is trained like this:

Average passes per image ≈ Training Steps × conversion factor ÷ number of images

Turn it upThe model learns the look more strongly, but may start copying your backgrounds and poses. Training time and credit cost go up too.

Turn it downTraining finishes sooner, but the model may not pick up the features you want.

Above 500 steps, the interface warns that training takes longer. Going past that is generally not recommended.

Learning rate

No effect on credits

learning_rate / unet_lr

6e-4 0.0006 · Default

PixAI Advanced Training learning rate slider at the default of 6.0e-4
Learning rate in the interface, default 6.0e-4

Learning rate controls how big an adjustment the model makes each time it learns. The default is 6e-4 (0.0006), and the range is 5e-5 (0.00005) to 1e-3 (0.001). Advanced Training has a single learning rate for the whole network, so you can't set separate rates per module as in kohya. Learning rate doesn't affect the credit cost.

Turn it upLearns faster, but may overshoot and give unstable results.

Turn it downLearns more carefully, but needs more steps.

Interface tip: lower it if results look messy or burnt.

Detail capacity (Rank)

No effect on credits

network_dim

PixAI Advanced Training detail capacity (Rank) options 8, 16, 32 and 64 with 64 as default
Detail capacity (Rank) in the interface, default 64

Detail capacity is kohya's network_dim: how much detail the LoRA can hold. Choose 8, 16, 32 or 64. The default is 64, and the interface labels the options Very light, Light, Balanced and Recommended. Rank mostly affects file size and how much detail is kept. It barely changes training time and doesn't affect the credit cost.

8Very light

16Light

32Balanced

64Recommended

Turn it upHolds finer detail, but memorizes your backgrounds sooner.

Turn it downSmaller file, but may miss complex patterns and accessories.

Alpha

No effect on credits

network_alpha

PixAI Advanced Training Alpha options with Rank 64: 32, 64 and 128, default 32 (Smoother)
Alpha in the interface: with Rank 64 you can pick 32, 64 or 128, default 32

Alpha is kohya's network_alpha. It sets how strongly each training update affects your LoRA. To keep it in step with Rank, Alpha can only be half, equal to, or double your Rank, shown in the interface as Smoother, Balanced and Sharper. The default is half your Rank. When you change Rank, Alpha keeps the same option and its value updates to match. Alpha doesn't affect the credit cost.

Turn it upLearns faster, but picks up flaws in your images just as fast.

Turn it downLearns more gently, which suits small image sets or mixed styles.

At Rank 64

32Smoother · ×0.5

64Balanced · ×1

128Sharper · ×2

The actual scaling comes from the ratio of Alpha to Rank. At Rank 64, the three options give Alpha values of 32, 64 and 128, a ratio of 0.5, 1 and 2. The common kohya setting of alpha = 1 gives a ratio of only 1/64 at Rank 64.

Update smoothing (Batch size)

Affects credits

batch_size

PixAI Advanced Training update smoothing (Batch size) options 1, 2 and 4 with 2 as default
Update smoothing (Batch size) in the interface, default 2

Update smoothing is kohya's batch size: how many images the model looks at before each update. Choose 1, 2 or 4. The default is 2. Larger values aren't available because of training cluster limits.

1

2Default

4

Turn it upSteadier training, and one odd image matters less. Each step processes twice as many images, so time and credit cost rise noticeably.

Turn it downSharper updates, but more sensitive to image quality.

Interface tip: leave it at 2 unless your images have noticeably different styles.

Batch size also changes how often each image is trained. At the same step count, going from 2 to 4 doubles the average passes per image and lengthens training. To keep the same amount of training, halve the steps.

Advanced Training vs. kohya / sd-scripts

In PixAICommon parameterDefault / optionsNotes
Training Stepsmax_train_steps325; 50–800 in steps of 5Counted as total steps, with no epochs or repeats. Every image has the same chance of being picked. Above 500 steps, the interface warns that training takes longer and fails more often.
Learning ratelearning_rate / unet_lr6e-4; 5e-5–1e-3One learning rate for the whole network.
Detail capacity (Rank)network_dim64; 8 / 16 / 32 / 64Mainly affects file size and how much detail is kept. Barely affects training time.
Alphanetwork_alphaHalf your Rank (32 at Rank 64); half, equal or doubleNo free values. Alpha doesn't affect the credit cost.
Update smoothing (Batch size)batch_size2; 1 / 2 / 4Higher values look at more images per update for steadier training. Larger values aren't available because of cluster limits.
Trigger wordstrigger word / class tokenRequired; letters a–z, numbers, spaces and , . _ - only; no minimum, 256 max; no consecutive, leading or trailing spacesEntered once in Basic Settings and added to the front of every description during training. Include them in your prompt to use the LoRA.
Image descriptions (captions)caption filesGenerated, then editableWritten in natural language by PixAI's description model, not Danbooru tags. Generate first, then edit. You can restore the generated version.

Credits and training time

Descriptions and training are billed separately in credits. Both show their cost before you confirm.

① Description cost · paid when you click Describe Automatically

100 × images + 500

Images that already have a description aren't charged again.

If it fails: only images described successfully are charged. Failed images can be retried.

② Training cost · paid when you click Start training

About 70,000 credits

Depends on your settings; about 70,000 credits with the defaults. Training Steps and batch size have the biggest effect. The quote shown before you start is the final price.

If it fails: training isn't charged, description costs aren't refunded. Use Try again with the same settings in Training tasks.

Two things differ from a kohya setup

Descriptions are a separate paid stage

Each run of automatic descriptions on undescribed images costs 100 × images described + 500 credits, shown before you confirm. Described images aren't refunded even if you delete them later.

Described images are never charged twice

Automatic descriptions only process images without one, so when you add images later you pay only for the new ones.

Advanced Training charges credits twice:

  1. Description cost: paid when you click Describe Automatically in the Descriptions step.
  2. Training cost: paid when you click Start training in the Advanced Parameters step, at the quoted price.

Examples

  • First trainingYou add 30 images and click Describe Automatically. You pay per image plus 500 credits, and the exact amount appears first. After checking descriptions and setting parameters, you click Start training and pay the quoted cost (about 70,000 credits with the defaults).
  • Adding imagesYou add 5 more images. Only the 5 new ones are described and charged per image, plus another 500 credits. The original 30 aren't charged again.
  • Some descriptions fail2 images can't be described. You pay for the 28 that succeeded, and the 500-credit fee still applies. Running automatic descriptions again on the 2 failed images adds another 500 credits.
  • Training failsThe training isn't charged, but description costs aren't refunded. Use Try again with the same settings in Training tasks.
  • CancellingCancel while queued and the training cost is refunded in full. Description costs aren't. The task becomes a draft you can edit and resubmit.

Training time

2–6 h excluding queue

Training time and credit cost depend mainly on Training Steps and batch size. Learning rate, Rank and Alpha don't change the cost and barely affect time. After you submit, the training waits in a queue first, so the finish time depends on how busy it is. See why LoRA training can take longer.

Advanced Training takes about 2 to 6 hours, not counting queue time. Your image set is locked while training runs and unlocks if the training fails or you cancel it in the queue. If training pauses, it resumes automatically at no extra cost.

Checklist before you start training

Before you click Start training, check these:

  • Images: at least 512×512 pixels in total, aspect ratio between 1:3 and 3:1, PNG, JPG or WebP, 10 to 100 images.
  • Image content: enough variety in angles, poses and framing, with no multi-panel images or near-duplicates. For a multi-character LoRA, each character alone plus images of them together.
  • Basic Settings: LoRA name, type, base model and trigger words confirmed. They can't be changed after you submit.
  • Descriptions: checked one by one, describing only visible things that vary, with the same word for the same thing. Every image has a description (Not described yet shows 0), or you can't continue.
  • Parameters: with no specific goal, keep the defaults (325 steps, learning rate 6e-4, Rank 64, Alpha 32, batch size 2).
  • Cost and time: you've checked the quoted cost and estimated time. You can only cancel while queued (training cost refunded, description cost not). Once training starts it can't be cancelled.

After training: publish and manage

When training finishes, click Publish in Training tasks. After that you can change the settings below any time from Edit Model on the model page.

Visibility

  • Private Hidden from others, and you can make it public later. Only members can keep LoRAs private, and there's a limit on how many; check your remaining slots under Edit Model. See Rule Update: Deleting Private LoRAs.
  • Public Anyone can use it. A public LoRA can't be made private again.

Rebates

Under Rebate settings in Edit Model, choose Opt in to rebate or Opt out of rebate. Once you opt in, you earn credits based on how often other users generate with your LoRA. Your own use doesn't count, so a private LoRA earns nothing even with rebates on.

Opting in can't be undone, and the LoRA can no longer be deleted. We recommend opting in only after the LoRA is public and you're sure you won't need to delete it.

Full details: LoRA Rebate Upgrade.

Downloads and commercial use

Tsubaki.2 and Tsubaki.3 are DiT models (shown as DiT.2 and DiT.3 on model pages). LoRAs trained on them can't be downloaded and can only be used on PixAI. See our DiT LoRA training guide.

Under Permissions in Edit Model, choose whether others may use images generated with this LoRA commercially.

Model details

You can change the model name, tags, description and cover any time in Edit Model. The LoRA type and base model can't be changed. Your training settings and final descriptions are saved with the model, so you can check them later and other users can find your LoRA more easily.

FAQ

Do I need to write the trigger words into every description?

No. They're added to the start of every description automatically. Enter them once in Basic Settings, and include them in your prompt when you generate.

Can I skip automatic descriptions and use my own captions?

Not at the moment. Every image needs a generated description before you can edit it, and importing existing caption files isn't supported.

Why are there no epochs or repeats?

Advanced Training counts total steps (Training Steps), and every image has the same chance of being picked. See Training Steps for how to convert.

Why can't I set Alpha to 1?

Alpha can only be half, equal to, or double your Rank. The actual scaling comes from the ratio of Alpha to Rank.

Why is batch size limited to 1, 2 or 4?

Training cluster limits. A larger batch size gives steadier updates but processes more images per step, which raises training time and credit cost. The default of 2 works in most cases.

Does Rank 64 slow training down a lot?

Barely, and it doesn't add to the credit cost. Time and cost depend mainly on Training Steps and batch size. Rank mostly affects file size and how much detail is kept.

Can I change the base model after training starts?

No. Once training starts, Basic Settings are locked and you'll need a new training. Before that you can still change them: go back before you submit, or cancel while queued to turn the task back into a draft.

Can I cancel after submitting?

Yes, while the training is queued. Once it starts, it can't be cancelled. Cancelling from Training tasks refunds the training cost in full but not description costs. The task becomes a draft, and you can edit the images, descriptions and Basic Settings (including the base model) before resubmitting.

Can I download my trained LoRA?

Not right now. Tsubaki.2 and Tsubaki.3 are DiT models, and LoRAs trained on them can only be used on PixAI.

To apply your LoRA and set its weight when generating, see how to use LoRA models and our LoRA weight settings guide.

Images ready?

Run through the checklist, then start your first Advanced Training on PixAI.