mirror of
https://github.com/invoke-ai/InvokeAI
synced 2024-08-30 20:32:17 +00:00
Merge branch 'main' into psyche/fix/ui/depth-anything-select
This commit is contained in:
blessedcoolant
GitHub
commit
cdc468a38c
@ -586,13 +586,6 @@ class DenoiseLatentsInvocation(BaseInvocation):
|
||||
unet: UNet2DConditionModel,
|
||||
scheduler: Scheduler,
|
||||
) -> StableDiffusionGeneratorPipeline:
|
||||
# TODO:
|
||||
# configure_model_padding(
|
||||
# unet,
|
||||
# self.seamless,
|
||||
# self.seamless_axes,
|
||||
# )
|
||||
|
||||
class FakeVae:
|
||||
class FakeVaeConfig:
|
||||
def __init__(self) -> None:
|
||||
|
||||
@ -4,5 +4,4 @@ Initialization file for invokeai.backend.image_util methods.
|
||||
|
||||
from .infill_methods.patchmatch import PatchMatch # noqa: F401
|
||||
from .pngwriter import PngWriter, PromptFormatter, retrieve_metadata, write_metadata # noqa: F401
|
||||
from .seamless import configure_model_padding # noqa: F401
|
||||
from .util import InitImageResizer, make_grid # noqa: F401
|
||||
|
||||
@ -1,52 +0,0 @@
|
||||
import torch.nn as nn
|
||||
|
||||
|
||||
def _conv_forward_asymmetric(self, input, weight, bias):
|
||||
"""
|
||||
Patch for Conv2d._conv_forward that supports asymmetric padding
|
||||
"""
|
||||
working = nn.functional.pad(input, self.asymmetric_padding["x"], mode=self.asymmetric_padding_mode["x"])
|
||||
working = nn.functional.pad(working, self.asymmetric_padding["y"], mode=self.asymmetric_padding_mode["y"])
|
||||
return nn.functional.conv2d(
|
||||
working,
|
||||
weight,
|
||||
bias,
|
||||
self.stride,
|
||||
nn.modules.utils._pair(0),
|
||||
self.dilation,
|
||||
self.groups,
|
||||
)
|
||||
|
||||
|
||||
def configure_model_padding(model, seamless, seamless_axes):
|
||||
"""
|
||||
Modifies the 2D convolution layers to use a circular padding mode based on
|
||||
the `seamless` and `seamless_axes` options.
|
||||
"""
|
||||
# TODO: get an explicit interface for this in diffusers: https://github.com/huggingface/diffusers/issues/556
|
||||
for m in model.modules():
|
||||
if isinstance(m, (nn.Conv2d, nn.ConvTranspose2d)):
|
||||
if seamless:
|
||||
m.asymmetric_padding_mode = {}
|
||||
m.asymmetric_padding = {}
|
||||
m.asymmetric_padding_mode["x"] = "circular" if ("x" in seamless_axes) else "constant"
|
||||
m.asymmetric_padding["x"] = (
|
||||
m._reversed_padding_repeated_twice[0],
|
||||
m._reversed_padding_repeated_twice[1],
|
||||
0,
|
||||
0,
|
||||
)
|
||||
m.asymmetric_padding_mode["y"] = "circular" if ("y" in seamless_axes) else "constant"
|
||||
m.asymmetric_padding["y"] = (
|
||||
0,
|
||||
0,
|
||||
m._reversed_padding_repeated_twice[2],
|
||||
m._reversed_padding_repeated_twice[3],
|
||||
)
|
||||
m._conv_forward = _conv_forward_asymmetric.__get__(m, nn.Conv2d)
|
||||
else:
|
||||
m._conv_forward = nn.Conv2d._conv_forward.__get__(m, nn.Conv2d)
|
||||
if hasattr(m, "asymmetric_padding_mode"):
|
||||
del m.asymmetric_padding_mode
|
||||
if hasattr(m, "asymmetric_padding"):
|
||||
del m.asymmetric_padding
|
||||
@ -1,89 +1,51 @@
|
||||
from __future__ import annotations
|
||||
|
||||
from contextlib import contextmanager
|
||||
from typing import Callable, List, Union
|
||||
from typing import Callable, List, Optional, Tuple, Union
|
||||
|
||||
import torch
|
||||
import torch.nn as nn
|
||||
from diffusers.models.autoencoders.autoencoder_kl import AutoencoderKL
|
||||
from diffusers.models.autoencoders.autoencoder_tiny import AutoencoderTiny
|
||||
from diffusers.models.lora import LoRACompatibleConv
|
||||
from diffusers.models.unets.unet_2d_condition import UNet2DConditionModel
|
||||
|
||||
|
||||
def _conv_forward_asymmetric(self, input, weight, bias):
|
||||
"""
|
||||
Patch for Conv2d._conv_forward that supports asymmetric padding
|
||||
"""
|
||||
working = nn.functional.pad(input, self.asymmetric_padding["x"], mode=self.asymmetric_padding_mode["x"])
|
||||
working = nn.functional.pad(working, self.asymmetric_padding["y"], mode=self.asymmetric_padding_mode["y"])
|
||||
return nn.functional.conv2d(
|
||||
working,
|
||||
weight,
|
||||
bias,
|
||||
self.stride,
|
||||
nn.modules.utils._pair(0),
|
||||
self.dilation,
|
||||
self.groups,
|
||||
)
|
||||
|
||||
|
||||
@contextmanager
|
||||
def set_seamless(model: Union[UNet2DConditionModel, AutoencoderKL, AutoencoderTiny], seamless_axes: List[str]):
|
||||
if not seamless_axes:
|
||||
yield
|
||||
return
|
||||
|
||||
# Callable: (input: Tensor, weight: Tensor, bias: Optional[Tensor]) -> Tensor
|
||||
to_restore: list[tuple[nn.Conv2d | nn.ConvTranspose2d, Callable]] = []
|
||||
# override conv_forward
|
||||
# https://github.com/huggingface/diffusers/issues/556#issuecomment-1993287019
|
||||
def _conv_forward_asymmetric(self, input: torch.Tensor, weight: torch.Tensor, bias: Optional[torch.Tensor] = None):
|
||||
self.paddingX = (self._reversed_padding_repeated_twice[0], self._reversed_padding_repeated_twice[1], 0, 0)
|
||||
self.paddingY = (0, 0, self._reversed_padding_repeated_twice[2], self._reversed_padding_repeated_twice[3])
|
||||
working = torch.nn.functional.pad(input, self.paddingX, mode=x_mode)
|
||||
working = torch.nn.functional.pad(working, self.paddingY, mode=y_mode)
|
||||
return torch.nn.functional.conv2d(
|
||||
working, weight, bias, self.stride, torch.nn.modules.utils._pair(0), self.dilation, self.groups
|
||||
)
|
||||
|
||||
original_layers: List[Tuple[nn.Conv2d, Callable]] = []
|
||||
|
||||
try:
|
||||
# Hard coded to skip down block layers, allowing for seamless tiling at the expense of prompt adherence
|
||||
skipped_layers = 1
|
||||
for m_name, m in model.named_modules():
|
||||
if not isinstance(m, (nn.Conv2d, nn.ConvTranspose2d)):
|
||||
continue
|
||||
x_mode = "circular" if "x" in seamless_axes else "constant"
|
||||
y_mode = "circular" if "y" in seamless_axes else "constant"
|
||||
|
||||
if isinstance(model, UNet2DConditionModel) and m_name.startswith("down_blocks.") and ".resnets." in m_name:
|
||||
# down_blocks.1.resnets.1.conv1
|
||||
_, block_num, _, resnet_num, submodule_name = m_name.split(".")
|
||||
block_num = int(block_num)
|
||||
resnet_num = int(resnet_num)
|
||||
conv_layers: List[torch.nn.Conv2d] = []
|
||||
|
||||
if block_num >= len(model.down_blocks) - skipped_layers:
|
||||
continue
|
||||
for module in model.modules():
|
||||
if isinstance(module, torch.nn.Conv2d):
|
||||
conv_layers.append(module)
|
||||
|
||||
# Skip the second resnet (could be configurable)
|
||||
if resnet_num > 0:
|
||||
continue
|
||||
|
||||
# Skip Conv2d layers (could be configurable)
|
||||
if submodule_name == "conv2":
|
||||
continue
|
||||
|
||||
m.asymmetric_padding_mode = {}
|
||||
m.asymmetric_padding = {}
|
||||
m.asymmetric_padding_mode["x"] = "circular" if ("x" in seamless_axes) else "constant"
|
||||
m.asymmetric_padding["x"] = (
|
||||
m._reversed_padding_repeated_twice[0],
|
||||
m._reversed_padding_repeated_twice[1],
|
||||
0,
|
||||
0,
|
||||
)
|
||||
m.asymmetric_padding_mode["y"] = "circular" if ("y" in seamless_axes) else "constant"
|
||||
m.asymmetric_padding["y"] = (
|
||||
0,
|
||||
0,
|
||||
m._reversed_padding_repeated_twice[2],
|
||||
m._reversed_padding_repeated_twice[3],
|
||||
)
|
||||
|
||||
to_restore.append((m, m._conv_forward))
|
||||
m._conv_forward = _conv_forward_asymmetric.__get__(m, nn.Conv2d)
|
||||
for layer in conv_layers:
|
||||
if isinstance(layer, LoRACompatibleConv) and layer.lora_layer is None:
|
||||
layer.lora_layer = lambda *x: 0
|
||||
original_layers.append((layer, layer._conv_forward))
|
||||
layer._conv_forward = _conv_forward_asymmetric.__get__(layer, torch.nn.Conv2d)
|
||||
|
||||
yield
|
||||
|
||||
finally:
|
||||
for module, orig_conv_forward in to_restore:
|
||||
module._conv_forward = orig_conv_forward
|
||||
if hasattr(module, "asymmetric_padding_mode"):
|
||||
del module.asymmetric_padding_mode
|
||||
if hasattr(module, "asymmetric_padding"):
|
||||
del module.asymmetric_padding
|
||||
for layer, orig_conv_forward in original_layers:
|
||||
layer._conv_forward = orig_conv_forward
|
||||
|
||||
@ -1,13 +1,14 @@
|
||||
import { isAnyOf } from '@reduxjs/toolkit';
|
||||
import { logger } from 'app/logging/logger';
|
||||
import type { AppStartListening } from 'app/store/middleware/listenerMiddleware';
|
||||
import type { AppDispatch } from 'app/store/store';
|
||||
import { parseify } from 'common/util/serialize';
|
||||
import {
|
||||
caLayerImageChanged,
|
||||
caLayerIsProcessingImageChanged,
|
||||
caLayerModelChanged,
|
||||
caLayerProcessedImageChanged,
|
||||
caLayerProcessorConfigChanged,
|
||||
caLayerProcessorPendingBatchIdChanged,
|
||||
caLayerRecalled,
|
||||
isControlAdapterLayer,
|
||||
} from 'features/controlLayers/store/controlLayersSlice';
|
||||
@ -15,47 +16,39 @@ import { CA_PROCESSOR_DATA } from 'features/controlLayers/util/controlAdapters';
|
||||
import { isImageOutput } from 'features/nodes/types/common';
|
||||
import { addToast } from 'features/system/store/systemSlice';
|
||||
import { t } from 'i18next';
|
||||
import { isEqual } from 'lodash-es';
|
||||
import { imagesApi } from 'services/api/endpoints/images';
|
||||
import { getImageDTO } from 'services/api/endpoints/images';
|
||||
import { queueApi } from 'services/api/endpoints/queue';
|
||||
import type { BatchConfig, ImageDTO } from 'services/api/types';
|
||||
import type { BatchConfig } from 'services/api/types';
|
||||
import { socketInvocationComplete } from 'services/events/actions';
|
||||
import { assert } from 'tsafe';
|
||||
|
||||
const matcher = isAnyOf(caLayerImageChanged, caLayerProcessorConfigChanged, caLayerModelChanged, caLayerRecalled);
|
||||
|
||||
const DEBOUNCE_MS = 300;
|
||||
const log = logger('session');
|
||||
|
||||
/**
|
||||
* Simple helper to cancel a batch and reset the pending batch ID
|
||||
*/
|
||||
const cancelProcessorBatch = async (dispatch: AppDispatch, layerId: string, batchId: string) => {
|
||||
const req = dispatch(queueApi.endpoints.cancelByBatchIds.initiate({ batch_ids: [batchId] }));
|
||||
log.trace({ batchId }, 'Cancelling existing preprocessor batch');
|
||||
try {
|
||||
await req.unwrap();
|
||||
} catch {
|
||||
// no-op
|
||||
} finally {
|
||||
req.reset();
|
||||
// Always reset the pending batch ID - the cancel req could fail if the batch doesn't exist
|
||||
dispatch(caLayerProcessorPendingBatchIdChanged({ layerId, batchId: null }));
|
||||
}
|
||||
};
|
||||
|
||||
export const addControlAdapterPreprocessor = (startAppListening: AppStartListening) => {
|
||||
startAppListening({
|
||||
matcher,
|
||||
effect: async (action, { dispatch, getState, getOriginalState, cancelActiveListeners, delay, take }) => {
|
||||
effect: async (action, { dispatch, getState, cancelActiveListeners, delay, take, signal }) => {
|
||||
const layerId = caLayerRecalled.match(action) ? action.payload.id : action.payload.layerId;
|
||||
const precheckLayerOriginal = getOriginalState()
|
||||
.controlLayers.present.layers.filter(isControlAdapterLayer)
|
||||
.find((l) => l.id === layerId);
|
||||
const precheckLayer = getState()
|
||||
.controlLayers.present.layers.filter(isControlAdapterLayer)
|
||||
.find((l) => l.id === layerId);
|
||||
|
||||
// Conditions to bail
|
||||
const layerDoesNotExist = !precheckLayer;
|
||||
const layerHasNoImage = !precheckLayer?.controlAdapter.image;
|
||||
const layerHasNoProcessorConfig = !precheckLayer?.controlAdapter.processorConfig;
|
||||
const layerIsAlreadyProcessingImage = precheckLayer?.controlAdapter.isProcessingImage;
|
||||
const areImageAndProcessorUnchanged =
|
||||
isEqual(precheckLayer?.controlAdapter.image, precheckLayerOriginal?.controlAdapter.image) &&
|
||||
isEqual(precheckLayer?.controlAdapter.processorConfig, precheckLayerOriginal?.controlAdapter.processorConfig);
|
||||
|
||||
if (
|
||||
layerDoesNotExist ||
|
||||
layerHasNoImage ||
|
||||
layerHasNoProcessorConfig ||
|
||||
areImageAndProcessorUnchanged ||
|
||||
layerIsAlreadyProcessingImage
|
||||
) {
|
||||
return;
|
||||
}
|
||||
|
||||
// Cancel any in-progress instances of this listener
|
||||
cancelActiveListeners();
|
||||
@ -63,19 +56,31 @@ export const addControlAdapterPreprocessor = (startAppListening: AppStartListeni
|
||||
|
||||
// Delay before starting actual work
|
||||
await delay(DEBOUNCE_MS);
|
||||
dispatch(caLayerIsProcessingImageChanged({ layerId, isProcessingImage: true }));
|
||||
|
||||
// Double-check that we are still eligible for processing
|
||||
const state = getState();
|
||||
const layer = state.controlLayers.present.layers.filter(isControlAdapterLayer).find((l) => l.id === layerId);
|
||||
const image = layer?.controlAdapter.image;
|
||||
const config = layer?.controlAdapter.processorConfig;
|
||||
|
||||
// If we have no image or there is no processor config, bail
|
||||
if (!layer || !image || !config) {
|
||||
if (!layer) {
|
||||
return;
|
||||
}
|
||||
|
||||
const image = layer.controlAdapter.image;
|
||||
const config = layer.controlAdapter.processorConfig;
|
||||
|
||||
if (!image || !config) {
|
||||
// The user has reset the image or config, so we should clear the processed image
|
||||
dispatch(caLayerProcessedImageChanged({ layerId, imageDTO: null }));
|
||||
}
|
||||
|
||||
// At this point, the user has stopped fiddling with the processor settings and there is a processor selected.
|
||||
|
||||
// If there is a pending processor batch, cancel it.
|
||||
if (layer.controlAdapter.processorPendingBatchId) {
|
||||
cancelProcessorBatch(dispatch, layerId, layer.controlAdapter.processorPendingBatchId);
|
||||
}
|
||||
|
||||
// @ts-expect-error: TS isn't able to narrow the typing of buildNode and `config` will error...
|
||||
const processorNode = CA_PROCESSOR_DATA[config.type].buildNode(image, config);
|
||||
const enqueueBatchArg: BatchConfig = {
|
||||
@ -83,7 +88,11 @@ export const addControlAdapterPreprocessor = (startAppListening: AppStartListeni
|
||||
batch: {
|
||||
graph: {
|
||||
nodes: {
|
||||
[processorNode.id]: { ...processorNode, is_intermediate: true },
|
||||
[processorNode.id]: {
|
||||
...processorNode,
|
||||
// Control images are always intermediate - do not save to gallery
|
||||
is_intermediate: true,
|
||||
},
|
||||
},
|
||||
edges: [],
|
||||
},
|
||||
@ -91,16 +100,21 @@ export const addControlAdapterPreprocessor = (startAppListening: AppStartListeni
|
||||
},
|
||||
};
|
||||
|
||||
try {
|
||||
// Kick off the processor batch
|
||||
const req = dispatch(
|
||||
queueApi.endpoints.enqueueBatch.initiate(enqueueBatchArg, {
|
||||
fixedCacheKey: 'enqueueBatch',
|
||||
})
|
||||
);
|
||||
|
||||
try {
|
||||
const enqueueResult = await req.unwrap();
|
||||
req.reset();
|
||||
// TODO(psyche): Update the pydantic models, pretty sure we will _always_ have a batch_id here, but the model says it's optional
|
||||
assert(enqueueResult.batch.batch_id, 'Batch ID not returned from queue');
|
||||
dispatch(caLayerProcessorPendingBatchIdChanged({ layerId, batchId: enqueueResult.batch.batch_id }));
|
||||
log.debug({ enqueueResult: parseify(enqueueResult) }, t('queue.graphQueued'));
|
||||
|
||||
// Wait for the processor node to complete
|
||||
const [invocationCompleteAction] = await take(
|
||||
(action): action is ReturnType<typeof socketInvocationComplete> =>
|
||||
socketInvocationComplete.match(action) &&
|
||||
@ -109,31 +123,33 @@ export const addControlAdapterPreprocessor = (startAppListening: AppStartListeni
|
||||
);
|
||||
|
||||
// We still have to check the output type
|
||||
if (isImageOutput(invocationCompleteAction.payload.data.result)) {
|
||||
assert(
|
||||
isImageOutput(invocationCompleteAction.payload.data.result),
|
||||
`Processor did not return an image output, got: ${invocationCompleteAction.payload.data.result}`
|
||||
);
|
||||
const { image_name } = invocationCompleteAction.payload.data.result.image;
|
||||
|
||||
// Wait for the ImageDTO to be received
|
||||
const [{ payload }] = await take(
|
||||
(action) =>
|
||||
imagesApi.endpoints.getImageDTO.matchFulfilled(action) && action.payload.image_name === image_name
|
||||
);
|
||||
|
||||
const imageDTO = payload as ImageDTO;
|
||||
const imageDTO = await getImageDTO(image_name);
|
||||
assert(imageDTO, "Failed to fetch processor output's image DTO");
|
||||
|
||||
// Whew! We made it. Update the layer with the processed image
|
||||
log.debug({ layerId, imageDTO }, 'ControlNet image processed');
|
||||
|
||||
// Update the processed image in the store
|
||||
dispatch(
|
||||
caLayerProcessedImageChanged({
|
||||
layerId,
|
||||
imageDTO,
|
||||
})
|
||||
);
|
||||
dispatch(caLayerIsProcessingImageChanged({ layerId, isProcessingImage: false }));
|
||||
}
|
||||
dispatch(caLayerProcessedImageChanged({ layerId, imageDTO }));
|
||||
dispatch(caLayerProcessorPendingBatchIdChanged({ layerId, batchId: null }));
|
||||
} catch (error) {
|
||||
if (signal.aborted) {
|
||||
// The listener was canceled - we need to cancel the pending processor batch, if there is one (could have changed by now).
|
||||
const pendingBatchId = getState()
|
||||
.controlLayers.present.layers.filter(isControlAdapterLayer)
|
||||
.find((l) => l.id === layerId)?.controlAdapter.processorPendingBatchId;
|
||||
if (pendingBatchId) {
|
||||
cancelProcessorBatch(dispatch, layerId, pendingBatchId);
|
||||
}
|
||||
log.trace('Control Adapter preprocessor cancelled');
|
||||
} else {
|
||||
// Some other error condition...
|
||||
console.log(error);
|
||||
log.error({ enqueueBatchArg: parseify(enqueueBatchArg) }, t('queue.graphFailedToQueue'));
|
||||
dispatch(caLayerIsProcessingImageChanged({ layerId, isProcessingImage: false }));
|
||||
|
||||
if (error instanceof Object) {
|
||||
if ('data' in error && 'status' in error) {
|
||||
@ -151,6 +167,9 @@ export const addControlAdapterPreprocessor = (startAppListening: AppStartListeni
|
||||
})
|
||||
);
|
||||
}
|
||||
} finally {
|
||||
req.reset();
|
||||
}
|
||||
},
|
||||
});
|
||||
};
|
||||
|
||||
@ -124,7 +124,7 @@ export const ControlAdapterImagePreview = memo(
|
||||
controlImage &&
|
||||
processedControlImage &&
|
||||
!isMouseOverImage &&
|
||||
!controlAdapter.isProcessingImage &&
|
||||
!controlAdapter.processorPendingBatchId &&
|
||||
controlAdapter.processorConfig !== null;
|
||||
|
||||
useEffect(() => {
|
||||
@ -190,7 +190,7 @@ export const ControlAdapterImagePreview = memo(
|
||||
/>
|
||||
</>
|
||||
|
||||
{controlAdapter.isProcessingImage && (
|
||||
{controlAdapter.processorPendingBatchId !== null && (
|
||||
<Flex
|
||||
position="absolute"
|
||||
top={0}
|
||||
|
||||
@ -27,7 +27,7 @@ import { modelChanged } from 'features/parameters/store/generationSlice';
|
||||
import type { ParameterAutoNegative } from 'features/parameters/types/parameterSchemas';
|
||||
import { getIsSizeOptimal, getOptimalDimension } from 'features/parameters/util/optimalDimension';
|
||||
import type { IRect, Vector2d } from 'konva/lib/types';
|
||||
import { isEqual, partition } from 'lodash-es';
|
||||
import { isEqual, partition, unset } from 'lodash-es';
|
||||
import { atom } from 'nanostores';
|
||||
import type { RgbColor } from 'react-colorful';
|
||||
import type { UndoableOptions } from 'redux-undo';
|
||||
@ -49,7 +49,7 @@ import type {
|
||||
} from './types';
|
||||
|
||||
export const initialControlLayersState: ControlLayersState = {
|
||||
_version: 2,
|
||||
_version: 3,
|
||||
selectedLayerId: null,
|
||||
brushSize: 100,
|
||||
layers: [],
|
||||
@ -334,13 +334,13 @@ export const controlLayersSlice = createSlice({
|
||||
const layer = selectCALayerOrThrow(state, layerId);
|
||||
layer.opacity = opacity;
|
||||
},
|
||||
caLayerIsProcessingImageChanged: (
|
||||
caLayerProcessorPendingBatchIdChanged: (
|
||||
state,
|
||||
action: PayloadAction<{ layerId: string; isProcessingImage: boolean }>
|
||||
action: PayloadAction<{ layerId: string; batchId: string | null }>
|
||||
) => {
|
||||
const { layerId, isProcessingImage } = action.payload;
|
||||
const { layerId, batchId } = action.payload;
|
||||
const layer = selectCALayerOrThrow(state, layerId);
|
||||
layer.controlAdapter.isProcessingImage = isProcessingImage;
|
||||
layer.controlAdapter.processorPendingBatchId = batchId;
|
||||
},
|
||||
//#endregion
|
||||
|
||||
@ -800,7 +800,7 @@ export const {
|
||||
caLayerProcessorConfigChanged,
|
||||
caLayerIsFilterEnabledChanged,
|
||||
caLayerOpacityChanged,
|
||||
caLayerIsProcessingImageChanged,
|
||||
caLayerProcessorPendingBatchIdChanged,
|
||||
// IPA Layers
|
||||
ipaLayerAdded,
|
||||
ipaLayerRecalled,
|
||||
@ -857,7 +857,16 @@ export const selectControlLayersSlice = (state: RootState) => state.controlLayer
|
||||
const migrateControlLayersState = (state: any): any => {
|
||||
if (state._version === 1) {
|
||||
// Reset state for users on v1 (e.g. beta users), some changes could cause
|
||||
return deepClone(initialControlLayersState);
|
||||
state = deepClone(initialControlLayersState);
|
||||
}
|
||||
if (state._version === 2) {
|
||||
// The CA `isProcessingImage` flag was replaced with a `processorPendingBatchId` property, fix up CA layers
|
||||
for (const layer of (state as ControlLayersState).layers) {
|
||||
if (layer.type === 'control_adapter_layer') {
|
||||
layer.controlAdapter.processorPendingBatchId = null;
|
||||
unset(layer.controlAdapter, 'isProcessingImage');
|
||||
}
|
||||
}
|
||||
}
|
||||
return state;
|
||||
};
|
||||
|
||||
@ -113,7 +113,7 @@ export const zLayer = z.discriminatedUnion('type', [
|
||||
export type Layer = z.infer<typeof zLayer>;
|
||||
|
||||
export type ControlLayersState = {
|
||||
_version: 2;
|
||||
_version: 3;
|
||||
selectedLayerId: string | null;
|
||||
layers: Layer[];
|
||||
brushSize: number;
|
||||
|
||||
@ -198,8 +198,8 @@ const zControlAdapterBase = z.object({
|
||||
weight: z.number().gte(0).lte(1),
|
||||
image: zImageWithDims.nullable(),
|
||||
processedImage: zImageWithDims.nullable(),
|
||||
isProcessingImage: z.boolean(),
|
||||
processorConfig: zProcessorConfig.nullable(),
|
||||
processorPendingBatchId: z.string().nullable().default(null),
|
||||
beginEndStepPct: zBeginEndStepPct,
|
||||
});
|
||||
|
||||
@ -521,8 +521,8 @@ export const initialControlNetV2: Omit<ControlNetConfigV2, 'id'> = {
|
||||
controlMode: 'balanced',
|
||||
image: null,
|
||||
processedImage: null,
|
||||
isProcessingImage: false,
|
||||
processorConfig: CA_PROCESSOR_DATA.canny_image_processor.buildDefaults(),
|
||||
processorPendingBatchId: null,
|
||||
};
|
||||
|
||||
export const initialT2IAdapterV2: Omit<T2IAdapterConfigV2, 'id'> = {
|
||||
@ -532,8 +532,8 @@ export const initialT2IAdapterV2: Omit<T2IAdapterConfigV2, 'id'> = {
|
||||
beginEndStepPct: [0, 1],
|
||||
image: null,
|
||||
processedImage: null,
|
||||
isProcessingImage: false,
|
||||
processorConfig: CA_PROCESSOR_DATA.canny_image_processor.buildDefaults(),
|
||||
processorPendingBatchId: null,
|
||||
};
|
||||
|
||||
export const initialIPAdapterV2: Omit<IPAdapterConfigV2, 'id'> = {
|
||||
|
||||
@ -587,7 +587,7 @@ const parseControlNetToControlAdapterLayer: MetadataParseFunc<ControlAdapterLaye
|
||||
image: imageDTO ? imageDTOToImageWithDims(imageDTO) : null,
|
||||
processedImage: processedImageDTO ? imageDTOToImageWithDims(processedImageDTO) : null,
|
||||
processorConfig,
|
||||
isProcessingImage: false,
|
||||
processorPendingBatchId: null,
|
||||
},
|
||||
};
|
||||
|
||||
@ -651,7 +651,7 @@ const parseT2IAdapterToControlAdapterLayer: MetadataParseFunc<ControlAdapterLaye
|
||||
image: imageDTO ? imageDTOToImageWithDims(imageDTO) : null,
|
||||
processedImage: processedImageDTO ? imageDTOToImageWithDims(processedImageDTO) : null,
|
||||
processorConfig,
|
||||
isProcessingImage: false,
|
||||
processorPendingBatchId: null,
|
||||
},
|
||||
};
|
||||
|
||||
|
||||
Loading…
Reference in New Issue
Block a user