towsty/aigen
1
0
Fork 0
Code Issues Pull Requests 1 Actions Packages Projects Releases Wiki Activity

Add Image Iterate: nested multi-prompt still families.

Browse Source 
Iterate runs a series of prompts as generate (no still), edit (still A), or compose (A+B), saves each output under one library collection like video extensions, and defaults to iterating from the original still.
This commit is contained in:
towsty
2026-09-04 22:46:19 -05:00
parent 279ac72049
commit 2435167a41
6 changed files with 261 additions and 130 deletions
Download Patch File Download Diff File Expand all files Collapse all files
server/api/v2/generate.post.ts
+38 -21
View File
@@ -131,7 +131,7 @@ export default defineEventHandler(async (event) => {
const mode = parseImageV2Mode(fields.mode)
if (!mode) {
throw createError({ statusCode: 400, statusMessage: 'mode must be edit, compose, refine, or generate' })
throw createError({ statusCode: 400, statusMessage: 'mode must be edit, compose, refine, generate, or iterate' })
}
const task = parseImageV2Task(
fields.task,
@@ -145,13 +145,19 @@ export default defineEventHandler(async (event) => {
const extraPasses = parsePasses(fields.passes)
const passMode = parseImagePassMode(fields.passMode)
const aspect = parseImageAspect(fields.aspect)
if (!promptMid && (mode === 'generate' || aspect === 'auto')) {
if (!promptMid && (mode === 'generate' || mode === 'iterate' || aspect === 'auto')) {
throw createError({ statusCode: 400, statusMessage: 'A prompt is required' })
}
if (extraPasses.length && (mode === 'refine' || mode === 'generate')) {
throw createError({
statusCode: 400,
statusMessage: 'Extra passes are only for Edit and Compose. Refine and Generate stay a single run.'
statusMessage: 'Extra passes are only for Edit, Compose, and Iterate. Refine and Generate stay a single run.'
})
}
if (mode === 'iterate' && !extraPasses.length) {
throw createError({
statusCode: 400,
statusMessage: 'Iterate needs at least two prompts. Add more iterations, or use Generate / Edit for a single run.'
})
}
if (extraPasses.some(item => !item.prompt)) {
@@ -172,13 +178,24 @@ export default defineEventHandler(async (event) => {
}
const ownerKey = libraryOwnerKey(event)
const imageA = mode === 'generate' ? null : await resolveImageRef(ownerKey, fields.image_a, uploadedA)
const imageB = mode === 'refine' || mode === 'generate' ? null : await resolveImageRef(ownerKey, fields.image_b, uploadedB)
const mask = mode === 'refine' ? await resolveImageRef(ownerKey, fields.mask, uploadedMask) : null
if (mode !== 'generate' && !imageA) {
const imageA = mode === 'generate'
? null
: await resolveImageRef(ownerKey, fields.image_a || fields.stillId, uploadedA).catch(() => null)
// Iterate with no still A is text-only generate iterations.
if (mode !== 'generate' && mode !== 'iterate' && !imageA) {
throw createError({ statusCode: 400, statusMessage: 'image_a is required' })
}
if (mode === 'iterate' && !imageA && (uploadedA || String(fields.image_a || fields.stillId || '').trim())) {
throw createError({ statusCode: 400, statusMessage: 'Could not load still A for iterate' })
}
const imageB = mode === 'refine' || mode === 'generate' || (mode === 'iterate' && !imageA)
? null
: await resolveImageRef(ownerKey, fields.image_b, uploadedB).catch(() => null)
const mask = mode === 'refine' ? await resolveImageRef(ownerKey, fields.mask, uploadedMask) : null
const graphMode = mode === 'iterate'
? (imageA ? (imageB ? 'compose' as const : 'edit' as const) : 'generate' as const)
: mode
if (mode === 'refine' && !mask) {
throw createError({
statusCode: 400,
@@ -220,7 +237,7 @@ export default defineEventHandler(async (event) => {
const cfg = engine === 'krea'
? clampImageCfg(fields.cfg, IMAGE_V2_KREA_CFG)
: turbo ? IMAGE_V2_TURBO_CFG : clampImageCfg(fields.cfg, IMAGE_V2_CFG_DEFAULT)
const megapixels = mode === 'generate'
const megapixels = graphMode === 'generate'
? 0
: fields.scaleToTotalPixels == null
? clampImageScaleMegapixels(fields.megapixels ?? fields.scaleMegapixels, 1)
@@ -230,31 +247,31 @@ export default defineEventHandler(async (event) => {
const seed = fields.seed && String(fields.seed) !== 'random'
? Number(fields.seed)
: Math.floor(Math.random() * 2_147_483_647)
const generateSize = mode === 'generate'
const generateSize = graphMode === 'generate'
? generateSizeForAspect(aspect)
: clampImageV2Size(fields.width, fields.height, fields.aspect)
const size = mode === 'generate'
const size = graphMode === 'generate'
? { width: generateSize.width, height: generateSize.height }
: imageDimensions(imageA!.data)
const v2Mode = mode as ImageV2Mode
const v2Task = (
mode === 'generate' ? 't2i' : mode === 'refine' ? 'refine' : mode === 'edit' ? 'scene' : task
graphMode === 'generate' ? 't2i' : graphMode === 'refine' ? 'refine' : graphMode === 'edit' ? 'scene' : task
) as ImageV2Task
const refineStrength = mode === 'generate'
const refineStrength = graphMode === 'generate'
? undefined
: engine === 'krea' || mode === 'refine'
: engine === 'krea' || graphMode === 'refine'
? clampImageV2Denoise(
fields.strength,
mode === 'compose' ? IMAGE_V2_KREA_COMPOSE_DENOISE : mode === 'edit' ? IMAGE_V2_KREA_EDIT_DENOISE : IMAGE_V2_DENOISE_DEFAULT
graphMode === 'compose' ? IMAGE_V2_KREA_COMPOSE_DENOISE : graphMode === 'edit' ? IMAGE_V2_KREA_EDIT_DENOISE : IMAGE_V2_DENOISE_DEFAULT
)
: undefined
const generateConsistencyModel = clampImageV2Strength(
fields.consistency_model,
mode === 'generate' || engine === 'krea' ? 0 : IMAGE_V2_CONSISTENCY_MODEL
graphMode === 'generate' || engine === 'krea' ? 0 : IMAGE_V2_CONSISTENCY_MODEL
)
const generateConsistencyClip = clampImageV2Strength(
fields.consistency_clip,
mode === 'generate' || engine === 'krea' ? 0 : IMAGE_V2_CONSISTENCY_CLIP
graphMode === 'generate' || engine === 'krea' ? 0 : IMAGE_V2_CONSISTENCY_CLIP
)
const conceptAllowed = isXaigenStudio()
const generateSnofsFallback = 0
@@ -325,7 +342,7 @@ export default defineEventHandler(async (event) => {
promptPost: promptPost || undefined,
name: clipName,
folderId,
aspect: mode === 'generate' ? generateSize.aspect : aspect,
aspect: graphMode === 'generate' ? generateSize.aspect : aspect,
width: size?.width || 0,
height: size?.height || 0,
steps,
@@ -349,10 +366,10 @@ export default defineEventHandler(async (event) => {
queueAutoRun: false,
negative: String(fields.negative || '').trim(),
passes: extraPasses,
passMode,
passMode: mode === 'iterate' && graphMode === 'generate' ? 'batch' : passMode,
referenceStillId: savedRef?.id,
referenceStillFilename: savedRef?.filename,
scaleToTotalPixels: mode !== 'generate' && megapixels > 0,
scaleToTotalPixels: graphMode !== 'generate' && megapixels > 0,
scaleMegapixels: megapixels,
imagePipeline: 'v2',
v2Mode,
@@ -386,7 +403,7 @@ export default defineEventHandler(async (event) => {
mode: v2Mode,
task: v2Task,
engine,
workflow: imageV2WorkflowFile(v2Mode, engine),
workflow: imageV2WorkflowFile(v2Mode, engine, { hasImage: Boolean(imageA), hasReference: Boolean(imageB) }),
width: size?.width || 0,
height: size?.height || 0,
strength: refineStrength,
server/utils/imageChainV2.ts
+37 -29
View File
@@ -11,6 +11,7 @@ import { comfyJobPrefix } from '~/utils/outputNames'
import { restorePromptParts, wrappedPromptForComfy } from '~/utils/promptParts'
import { isXaigenStudio } from '~/utils/videoModels'
import {
imageV2GraphMode,
imageV2StackSpecials,
mergeLegacyImageV2Loras,
parseImageV2Engine,
@@ -52,20 +53,24 @@ export type EditV2RunParams = {
export async function runEditV2(job: Job, params: EditV2RunParams) {
const library = job.library
if (!library) throw new Error('Edit job is missing library metadata')
if (params.mode === 'refine' && !params.mask) {
const graphMode = imageV2GraphMode(params.mode, {
hasImage: Boolean(params.image),
hasReference: Boolean(params.reference)
})
if (graphMode === 'refine' && !params.mask) {
throw new Error('Refine requires a mask. Refusing to fall back to Edit.')
}
if (params.mode === 'compose' && !params.reference) {
if (graphMode === 'compose' && !params.reference) {
throw new Error('Compose requires image B. Refusing to fall back to one-image generation.')
}
if (params.mode === 'edit' && params.reference) {
if (graphMode === 'edit' && params.reference && params.mode !== 'iterate') {
throw new Error('Edit mode takes one image. Use Compose for two stills.')
}
if (params.mode !== 'generate' && !params.image) {
if (graphMode !== 'generate' && !params.image) {
throw new Error('This v2 mode requires still A.')
}
const canChain = params.mode === 'edit' || params.mode === 'compose'
const canChain = params.mode === 'iterate' || params.mode === 'edit' || params.mode === 'compose'
const extraPasses = canChain
? (params.passes || []).map(item => ({ prompt: String(item.prompt || '').trim() })).filter(item => item.prompt)
: []
@@ -78,8 +83,9 @@ export async function runEditV2(job: Job, params: EditV2RunParams) {
...extraPasses.map(item => item.prompt)
]
const chainTotal = prompts.length
const batch = params.passMode !== 'chain'
const unit = batch ? 'Batch' : 'Pass'
// Text-only iterate is always independent generations. With a still, honor batch vs chain.
const batch = graphMode === 'generate' || params.passMode !== 'chain'
const unit = params.mode === 'iterate' ? 'Iteration' : (batch ? 'Batch' : 'Pass')
library.chainTotal = chainTotal
library.familyId = chainTotal > 1 ? (library.familyId || crypto.randomUUID()) : library.familyId
@@ -96,7 +102,7 @@ export async function runEditV2(job: Job, params: EditV2RunParams) {
job.imageComfyHost = getComfyHost()
await withImageComfyHost(job.imageComfyHost, async () => {
const generate = params.mode === 'generate'
const generate = graphMode === 'generate'
const engine = parseImageV2Engine(params.engine, 'flux')
const engineLabel = engine === 'krea' ? 'Krea' : 'Flux'
const { applyImageFill } = await import('~/server/utils/imagePad')
@@ -113,7 +119,7 @@ export async function runEditV2(job: Job, params: EditV2RunParams) {
const krea = engine === 'krea' ? await resolveKreaGenerateAssets() : null
const appliedStack = mergeLegacyImageV2Loras(params.loraStack?.length ? params.loraStack : library.loraStack, {
engine,
mode: params.mode,
mode: graphMode,
xaigen: isXaigenStudio(),
snofsModel: params.snofsModel,
snofsClip: params.snofsClip,
@@ -127,7 +133,7 @@ export async function runEditV2(job: Job, params: EditV2RunParams) {
const consistencyModel = appliedSpecials.consistency?.strengthModel ?? params.consistencyModel
const consistencyClip = appliedSpecials.consistency?.strengthClip ?? params.consistencyClip
if (!generate && params.mode === 'compose' && params.reference) {
if (!generate && graphMode === 'compose' && params.reference) {
uploadedRef = await uploadImage({
...params.reference,
filename: `ref_${params.reference.filename || 'image_b.png'}`
@@ -163,17 +169,17 @@ export async function runEditV2(job: Job, params: EditV2RunParams) {
emitChainJob(job, {
type: 'status',
message: generate
? `Queueing ${engineLabel} generate on Beast...`
: params.mode === 'refine'
? `Queueing ${engineLabel} iterate generate on Beast...`
: graphMode === 'refine'
? 'Uploading canvas and mask to Beast...'
: params.mode === 'compose' ? 'Uploading stills A and B to Beast...' : 'Uploading still A to Beast...',
: graphMode === 'compose' ? 'Uploading stills A and B to Beast...' : 'Uploading still A to Beast...',
progress: generate ? 12 : 8
})
if (index === 0 && !generate && image) {
const filled = await applyImageFill({
image,
mask: params.mode === 'refine' ? mask : null,
mask: graphMode === 'refine' ? mask : null,
aspect: params.aspect || 'auto'
})
image = filled.image
@@ -199,7 +205,7 @@ export async function runEditV2(job: Job, params: EditV2RunParams) {
prompt = composeFillPrompt(prompt, true)
}
if (index === 0 && params.mode === 'refine' && mask && !uploadedMask) {
if (index === 0 && graphMode === 'refine' && mask && !uploadedMask) {
uploadedMask = await uploadImage({
...mask,
filename: `mask_${mask.filename || 'refine-mask.png'}`
@@ -212,19 +218,21 @@ export async function runEditV2(job: Job, params: EditV2RunParams) {
if (!generate) {
emitChainJob(job, {
type: 'status',
message: params.mode === 'refine'
? `Queueing ${engineLabel} v2 refine on Beast...`
: params.mode === 'compose'
? `Queueing ${engineLabel} v2 compose (${params.task}) on Beast...`
: `Queueing ${engineLabel} v2 edit on Beast...`,
message: params.mode === 'iterate'
? `Queueing ${engineLabel} iterate (${graphMode}) on Beast...`
: graphMode === 'refine'
? `Queueing ${engineLabel} v2 refine on Beast...`
: graphMode === 'compose'
? `Queueing ${engineLabel} v2 compose (${params.task}) on Beast...`
: `Queueing ${engineLabel} v2 edit on Beast...`,
progress: 12
})
}
const built = buildImageV2Workflow({
mode: params.mode,
mode: graphMode,
engine,
task: params.task,
task: generate ? 't2i' : params.task,
prompt: wrappedPromptForComfy(library, prompt, index === 0 ? params.prompt : prompt),
negative: params.negative,
imageAName: uploaded?.name,
@@ -305,7 +313,7 @@ export async function runEditV2(job: Job, params: EditV2RunParams) {
kind: 'imagev2',
mode: params.mode,
engine,
task: params.task,
task: generate ? 't2i' : params.task,
negative: params.negative,
snofsModel,
snofsClip,
@@ -313,12 +321,12 @@ export async function runEditV2(job: Job, params: EditV2RunParams) {
consistencyClip,
steps: params.steps,
cfg: params.cfg,
megapixels: params.mode === 'generate' || !(params.megapixels > 0) ? undefined : params.megapixels,
scaleToTotalPixels: params.mode === 'generate' ? undefined : params.megapixels > 0,
megapixels: generate || !(params.megapixels > 0) ? undefined : params.megapixels,
scaleToTotalPixels: generate ? undefined : params.megapixels > 0,
turbo: params.turbo === true,
strength: params.mode === 'generate' ? undefined : params.strength,
width: params.mode === 'generate' ? params.width : undefined,
height: params.mode === 'generate' ? params.height : undefined,
strength: generate ? undefined : params.strength,
width: generate ? params.width : undefined,
height: generate ? params.height : undefined,
aspect: params.aspect || 'auto',
seed,
sourceStillId: params.sourceStillId || library.stillId,
@@ -362,7 +370,7 @@ export async function runEditV2(job: Job, params: EditV2RunParams) {
emitChainJob(job, {
type: 'complete',
message: chainTotal > 1
? `${engineLabel} v2 ${batch ? 'batch' : 'chain'} finished on Beast`
? `${engineLabel} ${params.mode === 'iterate' ? 'iterate' : `v2 ${batch ? 'batch' : 'chain'}`} finished on Beast`
: `${engineLabel} v2 finished on Beast`,
progress: 100,
stillId: still?.id,
server/utils/studioQueue.ts
+33 -26
View File
@@ -58,7 +58,7 @@ export interface StudioJobPayload {
scaleToTotalPixels?: boolean
scaleMegapixels?: number
imagePipeline?: 'v1' | 'v2'
v2Mode?: 'edit' | 'compose' | 'refine' | 'generate'
v2Mode?: 'edit' | 'compose' | 'refine' | 'generate' | 'iterate'
v2Task?: 'scene' | 'identity' | 'outfit' | 'face_lock' | 'refine' | 't2i'
engine?: 'flux' | 'krea'
snofsModel?: number
@@ -1017,7 +1017,8 @@ async function startStudioEditJob(item: StudioJob) {
const { runEditV2 } = await import('~/server/utils/imageChainV2')
const payload = item.payload
const generate = payload.imagePipeline === 'v2' && payload.v2Mode === 'generate'
const iterate = payload.imagePipeline === 'v2' && payload.v2Mode === 'iterate'
const generate = payload.imagePipeline === 'v2' && (payload.v2Mode === 'generate' || (iterate && !payload.stillId))
if (!generate && (!payload.stillId || !existsSync(stillPath(item.ownerKey, payload.stillId)))) {
throw new Error('The input still is missing from the library')
}
@@ -1038,30 +1039,36 @@ async function startStudioEditJob(item: StudioJob) {
: null
if (payload.imagePipeline === 'v2') {
const mode = payload.v2Mode === 'compose'
? 'compose'
: payload.v2Mode === 'refine'
? 'refine'
: payload.v2Mode === 'generate'
? 'generate'
: 'edit'
const mode = payload.v2Mode === 'iterate'
? 'iterate'
: payload.v2Mode === 'compose'
? 'compose'
: payload.v2Mode === 'refine'
? 'refine'
: payload.v2Mode === 'generate'
? 'generate'
: 'edit'
const graphMode = mode === 'iterate'
? (image ? (reference ? 'compose' : 'edit') : 'generate')
: mode
const maskId = payload.maskStillId
const mask = mode === 'refine' && maskId && existsSync(stillPath(item.ownerKey, maskId))
const mask = graphMode === 'refine' && maskId && existsSync(stillPath(item.ownerKey, maskId))
? {
filename: payload.maskStillFilename || 'refine-mask.png',
data: readFileSync(stillPath(item.ownerKey, maskId)),
type: 'image/png'
}
: null
if (mode === 'refine' && !mask) {
if (graphMode === 'refine' && !mask) {
throw new Error('Refine requires a mask. Refusing to fall back to Edit.')
}
if (mode === 'compose' && !reference) {
if (graphMode === 'compose' && !reference) {
throw new Error('Compose requires image B. Refusing to fall back to one-image generation.')
}
if (mode === 'edit' && reference) {
throw new Error('Edit mode takes one image. Use Compose for two stills.')
}
const passCount = (mode === 'edit' || mode === 'compose' || mode === 'iterate') ? (payload.passes?.length || 0) : 0
live = createEditLiveJob({
steps: payload.steps,
hideThumbnail: payload.hideThumbnail,
@@ -1088,9 +1095,9 @@ async function startStudioEditJob(item: StudioJob) {
familyId: item.familyId,
chainIndex: 0,
chainStep: 1,
chainTotal: 1 + ((mode === 'edit' || mode === 'compose') ? (payload.passes?.length || 0) : 0),
chainLabel: (mode === 'edit' || mode === 'compose') && payload.passes?.length
? (payload.passMode === 'chain' ? 'Pass 1' : 'Batch 1')
chainTotal: 1 + passCount,
chainLabel: passCount
? (mode === 'iterate' ? 'Iteration 1' : (payload.passMode === 'chain' ? 'Pass 1' : 'Batch 1'))
: undefined,
passes: payload.passes,
...persistLoraFields(payload.loraStack || payload.loraName)
@@ -1102,30 +1109,30 @@ async function startStudioEditJob(item: StudioJob) {
void runEditV2(live, {
mode,
engine: v2Engine,
task: payload.v2Task || 'scene',
image: mode === 'generate' ? null : image,
reference: mode === 'refine' || mode === 'generate' ? null : reference,
task: payload.v2Task || (graphMode === 'generate' ? 't2i' : 'scene'),
image: graphMode === 'generate' ? null : image,
reference: graphMode === 'compose' ? reference : null,
mask,
prompt: payload.prompt,
passes: payload.passes || [],
passMode: payload.passMode === 'chain' ? 'chain' : 'batch',
negative: payload.negative || '',
steps: payload.steps,
seed: live.library?.seed || payload.seed,
seed: payload.seed || Math.floor(Math.random() * 2_147_483_647),
cfg: payload.cfg,
snofsModel: v2Specials.snofs?.strengthModel ?? payload.snofsModel ?? 0,
snofsClip: v2Specials.snofs?.strengthClip ?? payload.snofsClip ?? 0,
consistencyModel: v2Specials.consistency?.strengthModel ?? payload.consistencyModel ?? (mode === 'generate' || payload.engine === 'krea' ? 0 : 0.7),
consistencyClip: v2Specials.consistency?.strengthClip ?? payload.consistencyClip ?? (mode === 'generate' || payload.engine === 'krea' ? 0 : 0.7),
megapixels: payload.scaleToTotalPixels === false ? 0 : (payload.scaleMegapixels ?? 1),
consistencyModel: v2Specials.consistency?.strengthModel ?? payload.consistencyModel ?? 0,
consistencyClip: v2Specials.consistency?.strengthClip ?? payload.consistencyClip ?? 0,
megapixels: payload.scaleMegapixels || 0,
strength: payload.refineStrength,
aspect: payload.aspect || 'auto',
width: payload.width,
height: payload.height,
turbo: payload.turbo === true,
aspect: payload.aspect,
sourceStillId: payload.stillId,
referenceStillId: payload.referenceStillId,
loraStack: payload.loraStack
loraStack: payload.loraStack,
passes: payload.passes,
passMode: payload.passMode === 'chain' ? 'chain' : 'batch',
}).catch((error) => {
const message = error instanceof Error ? error.message : String(error)
if (live && live.status !== 'error' && live.status !== 'cancelled' && live.status !== 'deferred') {