add gc after using consistency dec

Add support for SSD-1B

.github/ISSUE_TEMPLATE/bug_report.yml

@@ -1,25 +1,45 @@
 name: Bug Report
-description: You think somethings is broken in the UI
+description: You think something is broken in the UI
 title: "[Bug]: "
 labels: ["bug-report"]
 
 body:
-  - type: checkboxes
-    attributes:
-      label: Is there an existing issue for this?
-      description: Please search to see if an issue already exists for the bug you encountered, and that it hasn't been fixed in a recent build/commit.
-      options:
-        - label: I have searched the existing issues and checked the recent builds/commits
-          required: true
   - type: markdown
     attributes:
       value: |
-        *Please fill this form with as much information as possible, don't forget to fill "What OS..." and "What browsers" and *provide screenshots if possible**
+        > The title of the bug report should be short and descriptive.
+        > Use relevant keywords for searchability.
+        > Do not leave it blank, but also do not put an entire error log in it.
+  - type: checkboxes
+    attributes:
+      label: Checklist
+      description: |
+        Please perform basic debugging to see if extensions or configuration is the cause of the issue.
+        Basic debug procedure
+        　1. Disable all third-party extensions - check if extension is the cause
+        　2. Update extensions and webui - sometimes things just need to be updated
+        　3. Backup and remove your config.json and ui-config.json - check if the issue is caused by bad configuration
+        　4. Delete venv with third-party extensions disabled - sometimes extensions might cause wrong libraries to be installed
+        　5. Try a fresh installation webui in a different directory - see if a clean installation solves the issue
+        Before making a issue report please, check that the issue hasn't been reported recently.
+      options:
+        - label: The issue exists after disabling all extensions
+        - label: The issue exists on a clean installation of webui
+        - label: The issue is caused by an extension, but I believe it is caused by a bug in the webui
+        - label: The issue exists in the current version of the webui
+        - label: The issue has not been reported before recently
+        - label: The issue has been reported before but has not been fixed yet
+  - type: markdown
+    attributes:
+      value: |
+        > Please fill this form with as much information as possible. Don't forget to "Upload Sysinfo" and "What browsers" and provide screenshots if possible
   - type: textarea
     id: what-did
     attributes:
       label: What happened?
       description: Tell us what happened in a very clear and simple way
+      placeholder: |
+        txt2img is not working as intended.
     validations:
       required: true
   - type: textarea
@@ -27,9 +47,9 @@ body:
     attributes:
       label: Steps to reproduce the problem
       description: Please provide us with precise step by step instructions on how to reproduce the bug
-      value: |
-        1. Go to .... 
-        2. Press ....
+      placeholder: |
+        1. Go to ...
+        2. Press ...
         3. ...
     validations:
       required: true
@@ -38,13 +58,8 @@ body:
     attributes:
       label: What should have happened?
       description: Tell us what you think the normal behavior should be
-    validations:
-      required: true
-  - type: textarea
-    id: sysinfo
-    attributes:
-      label: Sysinfo
-      description: System info file, generated by WebUI. You can generate it in settings, on the Sysinfo page. Drag the file into the field to upload it. If you submit your report without including the sysinfo file, the report will be closed. If needed, review the report to make sure it includes no personal information you don't want to share. If you can't start WebUI, you can use --dump-sysinfo commandline argument to generate the file.
+      placeholder: |
+        WebUI should ...
     validations:
       required: true
   - type: dropdown
@@ -58,12 +73,25 @@ body:
         - Brave
         - Apple Safari
         - Microsoft Edge
+        - Android
+        - iOS
         - Other
+  - type: textarea
+    id: sysinfo
+    attributes:
+      label: Sysinfo
+      description: System info file, generated by WebUI. You can generate it in settings, on the Sysinfo page. Drag the file into the field to upload it. If you submit your report without including the sysinfo file, the report will be closed. If needed, review the report to make sure it includes no personal information you don't want to share. If you can't start WebUI, you can use --dump-sysinfo commandline argument to generate the file.
+      placeholder: |
+        1. Go to WebUI Settings -> Sysinfo -> Download system info.
+            If WebUI fails to launch, use --dump-sysinfo commandline argument to generate the file
+        2. Upload the Sysinfo as a attached file, Do NOT paste it in as plain text.
+    validations:
+      required: true
   - type: textarea
     id: logs
     attributes:
       label: Console logs
-      description: Please provide **full** cmd/terminal logs from the moment you started UI to the end of it, after your bug happened. If it's very long, provide a link to pastebin or similar service.
+      description: Please provide **full** cmd/terminal logs from the moment you started UI to the end of it, after the bug occured. If it's very long, provide a link to pastebin or similar service.
       render: Shell
     validations:
       required: true
@@ -71,4 +99,7 @@ body:
     id: misc
     attributes:
       label: Additional information
-      description: Please provide us with any relevant additional info or context.
+      description: | 
+        Please provide us with any relevant additional info or context.
+        Examples:
+        　I have updated my GPU driver recently.

CHANGELOG.md

@@ -1,3 +1,8 @@
+## 1.6.1
+
+### Bug Fixes:
+ * fix an error causing the webui to fail to start ([#13839](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/13839))
+
 ## 1.6.0
 
 ### Features:

README.md

@@ -91,6 +91,7 @@ A browser interface based on Gradio library for Stable Diffusion.
 - Eased resolution restriction: generated image's dimensions must be a multiple of 8 rather than 64
 - Now with a license!
 - Reorder elements in the UI from settings screen
+- [Segmind Stable Diffusion](https://huggingface.co/segmind/SSD-1B) support
 
 ## Installation and Running
 Make sure the required [dependencies](https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Dependencies) are met and follow the instructions available for:

configs/alt-diffusion-m18-inference.yaml

@@ -0,0 +1,73 @@
+model:
+  base_learning_rate: 1.0e-04
+  target: ldm.models.diffusion.ddpm.LatentDiffusion
+  params:
+    linear_start: 0.00085
+    linear_end: 0.0120
+    num_timesteps_cond: 1
+    log_every_t: 200
+    timesteps: 1000
+    first_stage_key: "jpg"
+    cond_stage_key: "txt"
+    image_size: 64
+    channels: 4
+    cond_stage_trainable: false   # Note: different from the one we trained before
+    conditioning_key: crossattn
+    monitor: val/loss_simple_ema
+    scale_factor: 0.18215
+    use_ema: False
+
+    scheduler_config: # 10000 warmup steps
+      target: ldm.lr_scheduler.LambdaLinearScheduler
+      params:
+        warm_up_steps: [ 10000 ]
+        cycle_lengths: [ 10000000000000 ] # incredibly large number to prevent corner cases
+        f_start: [ 1.e-6 ]
+        f_max: [ 1. ]
+        f_min: [ 1. ]
+
+    unet_config:
+      target: ldm.modules.diffusionmodules.openaimodel.UNetModel
+      params:
+        image_size: 32 # unused
+        in_channels: 4
+        out_channels: 4
+        model_channels: 320
+        attention_resolutions: [ 4, 2, 1 ]
+        num_res_blocks: 2
+        channel_mult: [ 1, 2, 4, 4 ]
+        num_head_channels: 64
+        use_spatial_transformer: True
+        use_linear_in_transformer: True
+        transformer_depth: 1
+        context_dim: 1024
+        use_checkpoint: True
+        legacy: False
+
+    first_stage_config:
+      target: ldm.models.autoencoder.AutoencoderKL
+      params:
+        embed_dim: 4
+        monitor: val/rec_loss
+        ddconfig:
+          double_z: true
+          z_channels: 4
+          resolution: 256
+          in_channels: 3
+          out_ch: 3
+          ch: 128
+          ch_mult:
+          - 1
+          - 2
+          - 4
+          - 4
+          num_res_blocks: 2
+          attn_resolutions: []
+          dropout: 0.0
+        lossconfig:
+          target: torch.nn.Identity
+
+    cond_stage_config:
+      target: modules.xlmr_m18.BertSeriesModelWithTransformation
+      params:
+        name: "XLMR-Large"

extensions-builtin/Lora/lora_logger.py

@@ -0,0 +1,33 @@
+import sys
+import copy
+import logging
+
+
+class ColoredFormatter(logging.Formatter):
+    COLORS = {
+        "DEBUG": "\033[0;36m",  # CYAN
+        "INFO": "\033[0;32m",  # GREEN
+        "WARNING": "\033[0;33m",  # YELLOW
+        "ERROR": "\033[0;31m",  # RED
+        "CRITICAL": "\033[0;37;41m",  # WHITE ON RED
+        "RESET": "\033[0m",  # RESET COLOR
+    }
+
+    def format(self, record):
+        colored_record = copy.copy(record)
+        levelname = colored_record.levelname
+        seq = self.COLORS.get(levelname, self.COLORS["RESET"])
+        colored_record.levelname = f"{seq}{levelname}{self.COLORS['RESET']}"
+        return super().format(colored_record)
+
+
+logger = logging.getLogger("lora")
+logger.propagate = False
+
+
+if not logger.handlers:
+    handler = logging.StreamHandler(sys.stdout)
+    handler.setFormatter(
+        ColoredFormatter("[%(name)s]-%(levelname)s: %(message)s")
+    )
+    logger.addHandler(handler)

extensions-builtin/Lora/network.py

@@ -93,6 +93,7 @@ class Network:  # LoraModule
         self.unet_multiplier = 1.0
         self.dyn_dim = None
         self.modules = {}
+        self.bundle_embeddings = {}
         self.mtime = None
 
         self.mentioned_name = None

extensions-builtin/Lora/network_glora.py

@@ -0,0 +1,33 @@
+
+import network
+
+class ModuleTypeGLora(network.ModuleType):
+    def create_module(self, net: network.Network, weights: network.NetworkWeights):
+        if all(x in weights.w for x in ["a1.weight", "a2.weight", "alpha", "b1.weight", "b2.weight"]):
+            return NetworkModuleGLora(net, weights)
+
+        return None
+
+# adapted from https://github.com/KohakuBlueleaf/LyCORIS
+class NetworkModuleGLora(network.NetworkModule):
+    def __init__(self,  net: network.Network, weights: network.NetworkWeights):
+        super().__init__(net, weights)
+
+        if hasattr(self.sd_module, 'weight'):
+            self.shape = self.sd_module.weight.shape
+
+        self.w1a = weights.w["a1.weight"]
+        self.w1b = weights.w["b1.weight"]
+        self.w2a = weights.w["a2.weight"]
+        self.w2b = weights.w["b2.weight"]
+
+    def calc_updown(self, orig_weight):
+        w1a = self.w1a.to(orig_weight.device, dtype=orig_weight.dtype)
+        w1b = self.w1b.to(orig_weight.device, dtype=orig_weight.dtype)
+        w2a = self.w2a.to(orig_weight.device, dtype=orig_weight.dtype)
+        w2b = self.w2b.to(orig_weight.device, dtype=orig_weight.dtype)
+
+        output_shape = [w1a.size(0), w1b.size(1)]
+        updown = ((w2b @ w1b) + ((orig_weight @ w2a) @ w1a))
+
+        return self.finalize_updown(updown, orig_weight, output_shape)

extensions-builtin/Lora/networks.py

@@ -5,6 +5,7 @@ import re
 import lora_patches
 import network
 import network_lora
+import network_glora
 import network_hada
 import network_ia3
 import network_lokr
@@ -15,6 +16,9 @@ import torch
 from typing import Union
 
 from modules import shared, devices, sd_models, errors, scripts, sd_hijack
+import modules.textual_inversion.textual_inversion as textual_inversion
+
+from lora_logger import logger
 
 module_types = [
     network_lora.ModuleTypeLora(),
@@ -23,6 +27,7 @@ module_types = [
     network_lokr.ModuleTypeLokr(),
     network_full.ModuleTypeFull(),
     network_norm.ModuleTypeNorm(),
+    network_glora.ModuleTypeGLora(),
 ]
 
 
@@ -149,9 +154,19 @@ def load_network(name, network_on_disk):
     is_sd2 = 'model_transformer_resblocks' in shared.sd_model.network_layer_mapping
 
     matched_networks = {}
+    bundle_embeddings = {}
 
     for key_network, weight in sd.items():
         key_network_without_network_parts, network_part = key_network.split(".", 1)
+        if key_network_without_network_parts == "bundle_emb":
+            emb_name, vec_name = network_part.split(".", 1)
+            emb_dict = bundle_embeddings.get(emb_name, {})
+            if vec_name.split('.')[0] == 'string_to_param':
+                _, k2 = vec_name.split('.', 1)
+                emb_dict['string_to_param'] = {k2: weight}
+            else:
+                emb_dict[vec_name] = weight
+            bundle_embeddings[emb_name] = emb_dict
 
         key = convert_diffusers_name_to_compvis(key_network_without_network_parts, is_sd2)
         sd_module = shared.sd_model.network_layer_mapping.get(key, None)
@@ -195,6 +210,14 @@ def load_network(name, network_on_disk):
 
         net.modules[key] = net_module
 
+    embeddings = {}
+    for emb_name, data in bundle_embeddings.items():
+        embedding = textual_inversion.create_embedding_from_data(data, emb_name, filename=network_on_disk.filename + "/" + emb_name)
+        embedding.loaded = None
+        embeddings[emb_name] = embedding
+
+    net.bundle_embeddings = embeddings
+
     if keys_failed_to_match:
         logging.debug(f"Network {network_on_disk.filename} didn't match keys: {keys_failed_to_match}")
 
@@ -210,11 +233,15 @@ def purge_networks_from_memory():
 
 
 def load_networks(names, te_multipliers=None, unet_multipliers=None, dyn_dims=None):
+    emb_db = sd_hijack.model_hijack.embedding_db
     already_loaded = {}
 
     for net in loaded_networks:
         if net.name in names:
             already_loaded[net.name] = net
+        for emb_name, embedding in net.bundle_embeddings.items():
+            if embedding.loaded:
+                emb_db.register_embedding_by_name(None, shared.sd_model, emb_name)
 
     loaded_networks.clear()
 
@@ -257,6 +284,21 @@ def load_networks(names, te_multipliers=None, unet_multipliers=None, dyn_dims=No
         net.dyn_dim = dyn_dims[i] if dyn_dims else 1.0
         loaded_networks.append(net)
 
+        for emb_name, embedding in net.bundle_embeddings.items():
+            if embedding.loaded is None and emb_name in emb_db.word_embeddings:
+                logger.warning(
+                    f'Skip bundle embedding: "{emb_name}"'
+                    ' as it was already loaded from embeddings folder'
+                )
+                continue
+
+            embedding.loaded = False
+            if emb_db.expected_shape == -1 or emb_db.expected_shape == embedding.shape:
+                embedding.loaded = True
+                emb_db.register_embedding(embedding, shared.sd_model)
+            else:
+                emb_db.skipped_embeddings[name] = embedding
+
     if failed_to_load_networks:
         sd_hijack.model_hijack.comments.append("Networks not found: " + ", ".join(failed_to_load_networks))
 
@@ -565,6 +607,7 @@ extra_network_lora = None
 available_networks = {}
 available_network_aliases = {}
 loaded_networks = []
+loaded_bundle_embeddings = {}
 networks_in_memory = {}
 available_network_hash_lookup = {}
 forbidden_network_aliases = {}

extensions-builtin/mobile/javascript/mobile.js

@@ -12,6 +12,8 @@ function isMobile() {
 }
 
 function reportWindowSize() {
+    if (gradioApp().querySelector('.toprow-compact-tools')) return; // not applicable for compact prompt layout
+
     var currentlyMobile = isMobile();
     if (currentlyMobile == isSetupForMobile) return;
     isSetupForMobile = currentlyMobile;

javascript/edit-attention.js

@@ -19,16 +19,28 @@ function keyupEditAttention(event) {
         let beforeParen = before.lastIndexOf(OPEN);
         if (beforeParen == -1) return false;
 
+        let beforeClosingParen = before.lastIndexOf(CLOSE);
+        if (beforeClosingParen != -1 && beforeClosingParen > beforeParen) return false;
+
         // Find closing parenthesis around current cursor
         const after = text.substring(selectionStart);
         let afterParen = after.indexOf(CLOSE);
         if (afterParen == -1) return false;
 
+        let afterOpeningParen = after.indexOf(OPEN);
+        if (afterOpeningParen != -1 && afterOpeningParen < afterParen) return false;
+
         // Set the selection to the text between the parenthesis
         const parenContent = text.substring(beforeParen + 1, selectionStart + afterParen);
-        const lastColon = parenContent.lastIndexOf(":");
-        selectionStart = beforeParen + 1;
-        selectionEnd = selectionStart + lastColon;
+        if (/.*:-?[\d.]+/s.test(parenContent)) {
+            const lastColon = parenContent.lastIndexOf(":");
+            selectionStart = beforeParen + 1;
+            selectionEnd = selectionStart + lastColon;
+        } else {
+            selectionStart = beforeParen + 1;
+            selectionEnd = selectionStart + parenContent.length;
+        }
+
         target.setSelectionRange(selectionStart, selectionEnd);
         return true;
     }
@@ -57,7 +69,7 @@ function keyupEditAttention(event) {
     }
 
     // If the user hasn't selected anything, let's select their current parenthesis block or word
-    if (!selectCurrentParenthesisBlock('<', '>') && !selectCurrentParenthesisBlock('(', ')')) {
+    if (!selectCurrentParenthesisBlock('<', '>') && !selectCurrentParenthesisBlock('(', ')') && !selectCurrentParenthesisBlock('[', ']')) {
         selectCurrentWord();
     }
 
@@ -65,33 +77,54 @@ function keyupEditAttention(event) {
 
     var closeCharacter = ')';
     var delta = opts.keyedit_precision_attention;
+    var start = selectionStart > 0 ? text[selectionStart - 1] : "";
+    var end = text[selectionEnd];
 
-    if (selectionStart > 0 && text[selectionStart - 1] == '<') {
+    if (start == '<') {
         closeCharacter = '>';
         delta = opts.keyedit_precision_extra;
-    } else if (selectionStart == 0 || text[selectionStart - 1] != "(") {
+    } else if (start == '(' && end == ')' || start == '[' && end == ']') { // convert old-style (((emphasis)))
+        let numParen = 0;
 
+        while (text[selectionStart - numParen - 1] == start && text[selectionEnd + numParen] == end) {
+            numParen++;
+        }
+
+        if (start == "[") {
+            weight = (1 / 1.1) ** numParen;
+        } else {
+            weight = 1.1 ** numParen;
+        }
+
+        weight = Math.round(weight / opts.keyedit_precision_attention) * opts.keyedit_precision_attention;
+
+        text = text.slice(0, selectionStart - numParen) + "(" + text.slice(selectionStart, selectionEnd) + ":" + weight + ")" + text.slice(selectionEnd + numParen);
+        selectionStart -= numParen - 1;
+        selectionEnd -= numParen - 1;
+    } else if (start != '(') {
         // do not include spaces at the end
         while (selectionEnd > selectionStart && text[selectionEnd - 1] == ' ') {
-            selectionEnd -= 1;
+            selectionEnd--;
         }
+
         if (selectionStart == selectionEnd) {
             return;
         }
 
         text = text.slice(0, selectionStart) + "(" + text.slice(selectionStart, selectionEnd) + ":1.0)" + text.slice(selectionEnd);
 
-        selectionStart += 1;
-        selectionEnd += 1;
+        selectionStart++;
+        selectionEnd++;
     }
 
-    var end = text.slice(selectionEnd + 1).indexOf(closeCharacter) + 1;
-    var weight = parseFloat(text.slice(selectionEnd + 1, selectionEnd + end));
+    if (text[selectionEnd] != ':') return;
+    var weightLength = text.slice(selectionEnd + 1).indexOf(closeCharacter) + 1;
+    var weight = parseFloat(text.slice(selectionEnd + 1, selectionEnd + weightLength));
     if (isNaN(weight)) return;
 
     weight += isPlus ? delta : -delta;
     weight = parseFloat(weight.toPrecision(12));
-    if (String(weight).length == 1) weight += ".0";
+    if (Number.isInteger(weight)) weight += ".0";
 
     if (closeCharacter == ')' && weight == 1) {
         var endParenPos = text.substring(selectionEnd).indexOf(')');
@@ -99,7 +132,7 @@ function keyupEditAttention(event) {
         selectionStart--;
         selectionEnd--;
     } else {
-        text = text.slice(0, selectionEnd + 1) + weight + text.slice(selectionEnd + end);
+        text = text.slice(0, selectionEnd + 1) + weight + text.slice(selectionEnd + weightLength);
     }
 
     target.focus();

javascript/extraNetworks.js

@@ -26,8 +26,9 @@ function setupExtraNetworksForTab(tabname) {
     var refresh = gradioApp().getElementById(tabname + '_extra_refresh');
     var showDirsDiv = gradioApp().getElementById(tabname + '_extra_show_dirs');
     var showDirs = gradioApp().querySelector('#' + tabname + '_extra_show_dirs input');
+    var promptContainer = gradioApp().querySelector('.prompt-container-compact#' + tabname + '_prompt_container');
+    var negativePrompt = gradioApp().querySelector('#' + tabname + '_neg_prompt');
 
-    sort.dataset.sortkey = 'sortDefault';
     tabs.appendChild(searchDiv);
     tabs.appendChild(sort);
     tabs.appendChild(sortOrder);
@@ -49,20 +50,23 @@ function setupExtraNetworksForTab(tabname) {
 
             elem.style.display = visible ? "" : "none";
         });
+
+        applySort();
     };
 
     var applySort = function() {
+        var cards = gradioApp().querySelectorAll('#' + tabname + '_extra_tabs div.card');
+
         var reverse = sortOrder.classList.contains("sortReverse");
-        var sortKey = sort.querySelector("input").value.toLowerCase().replace("sort", "").replaceAll(" ", "_").replace(/_+$/, "").trim();
-        sortKey = sortKey ? "sort" + sortKey.charAt(0).toUpperCase() + sortKey.slice(1) : "";
-        var sortKeyStore = sortKey ? sortKey + (reverse ? "Reverse" : "") : "";
-        if (!sortKey || sortKeyStore == sort.dataset.sortkey) {
+        var sortKey = sort.querySelector("input").value.toLowerCase().replace("sort", "").replaceAll(" ", "_").replace(/_+$/, "").trim() || "name";
+        sortKey = "sort" + sortKey.charAt(0).toUpperCase() + sortKey.slice(1);
+        var sortKeyStore = sortKey + "-" + (reverse ? "Descending" : "Ascending") + "-" + cards.length;
+
+        if (sortKeyStore == sort.dataset.sortkey) {
             return;
         }
-
         sort.dataset.sortkey = sortKeyStore;
 
-        var cards = gradioApp().querySelectorAll('#' + tabname + '_extra_tabs div.card');
         cards.forEach(function(card) {
             card.originalParentElement = card.parentElement;
         });
@@ -88,15 +92,13 @@ function setupExtraNetworksForTab(tabname) {
     };
 
     search.addEventListener("input", applyFilter);
-    applyFilter();
-    ["change", "blur", "click"].forEach(function(evt) {
-        sort.querySelector("input").addEventListener(evt, applySort);
-    });
     sortOrder.addEventListener("click", function() {
         sortOrder.classList.toggle("sortReverse");
         applySort();
     });
+    applyFilter();
 
+    extraNetworksApplySort[tabname] = applySort;
     extraNetworksApplyFilter[tabname] = applyFilter;
 
     var showDirsUpdate = function() {
@@ -109,11 +111,47 @@ function setupExtraNetworksForTab(tabname) {
     showDirsUpdate();
 }
 
+function extraNetworksMovePromptToTab(tabname, id, showPrompt, showNegativePrompt) {
+    if (!gradioApp().querySelector('.toprow-compact-tools')) return; // only applicable for compact prompt layout
+
+    var promptContainer = gradioApp().getElementById(tabname + '_prompt_container');
+    var prompt = gradioApp().getElementById(tabname + '_prompt_row');
+    var negPrompt = gradioApp().getElementById(tabname + '_neg_prompt_row');
+    var elem = id ? gradioApp().getElementById(id) : null;
+
+    if (showNegativePrompt && elem) {
+        elem.insertBefore(negPrompt, elem.firstChild);
+    } else {
+        promptContainer.insertBefore(negPrompt, promptContainer.firstChild);
+    }
+
+    if (showPrompt && elem) {
+        elem.insertBefore(prompt, elem.firstChild);
+    } else {
+        promptContainer.insertBefore(prompt, promptContainer.firstChild);
+    }
+}
+
+
+function extraNetworksUrelatedTabSelected(tabname) { // called from python when user selects an unrelated tab (generate)
+    extraNetworksMovePromptToTab(tabname, '', false, false);
+}
+
+function extraNetworksTabSelected(tabname, id, showPrompt, showNegativePrompt) { // called from python when user selects an extra networks tab
+    extraNetworksMovePromptToTab(tabname, id, showPrompt, showNegativePrompt);
+
+}
+
 function applyExtraNetworkFilter(tabname) {
     setTimeout(extraNetworksApplyFilter[tabname], 1);
 }
 
+function applyExtraNetworkSort(tabname) {
+    setTimeout(extraNetworksApplySort[tabname], 1);
+}
+
 var extraNetworksApplyFilter = {};
+var extraNetworksApplySort = {};
 var activePromptTextarea = {};
 
 function setupExtraNetworks() {

javascript/imageviewer.js

@@ -33,8 +33,11 @@ function updateOnBackgroundChange() {
     const modalImage = gradioApp().getElementById("modalImage");
     if (modalImage && modalImage.offsetParent) {
         let currentButton = selected_gallery_button();
-
-        if (currentButton?.children?.length > 0 && modalImage.src != currentButton.children[0].src) {
+        let preview = gradioApp().querySelectorAll('.livePreview > img');
+        if (preview.length > 0) {
+            // show preview image if available
+            modalImage.src = preview[preview.length - 1].src;
+        } else if (currentButton?.children?.length > 0 && modalImage.src != currentButton.children[0].src) {
             modalImage.src = currentButton.children[0].src;
             if (modalImage.style.display === 'none') {
                 const modal = gradioApp().getElementById("lightboxModal");

javascript/inputAccordion.js

@@ -1,37 +1,68 @@
-var observerAccordionOpen = new MutationObserver(function(mutations) {
-    mutations.forEach(function(mutationRecord) {
-        var elem = mutationRecord.target;
-        var open = elem.classList.contains('open');
-
-        var accordion = elem.parentNode;
-        accordion.classList.toggle('input-accordion-open', open);
-
-        var checkbox = gradioApp().querySelector('#' + accordion.id + "-checkbox input");
-        checkbox.checked = open;
-        updateInput(checkbox);
-
-        var extra = gradioApp().querySelector('#' + accordion.id + "-extra");
-        if (extra) {
-            extra.style.display = open ? "" : "none";
-        }
-    });
-});
-
 function inputAccordionChecked(id, checked) {
-    var label = gradioApp().querySelector('#' + id + " .label-wrap");
-    if (label.classList.contains('open') != checked) {
-        label.click();
+    var accordion = gradioApp().getElementById(id);
+    accordion.visibleCheckbox.checked = checked;
+    accordion.onVisibleCheckboxChange();
+}
+
+function setupAccordion(accordion) {
+    var labelWrap = accordion.querySelector('.label-wrap');
+    var gradioCheckbox = gradioApp().querySelector('#' + accordion.id + "-checkbox input");
+    var extra = gradioApp().querySelector('#' + accordion.id + "-extra");
+    var span = labelWrap.querySelector('span');
+    var linked = true;
+
+    var isOpen = function() {
+        return labelWrap.classList.contains('open');
+    };
+
+    var observerAccordionOpen = new MutationObserver(function(mutations) {
+        mutations.forEach(function(mutationRecord) {
+            accordion.classList.toggle('input-accordion-open', isOpen());
+
+            if (linked) {
+                accordion.visibleCheckbox.checked = isOpen();
+                accordion.onVisibleCheckboxChange();
+            }
+        });
+    });
+    observerAccordionOpen.observe(labelWrap, {attributes: true, attributeFilter: ['class']});
+
+    if (extra) {
+        labelWrap.insertBefore(extra, labelWrap.lastElementChild);
     }
+
+    accordion.onChecked = function(checked) {
+        if (isOpen() != checked) {
+            labelWrap.click();
+        }
+    };
+
+    var visibleCheckbox = document.createElement('INPUT');
+    visibleCheckbox.type = 'checkbox';
+    visibleCheckbox.checked = isOpen();
+    visibleCheckbox.id = accordion.id + "-visible-checkbox";
+    visibleCheckbox.className = gradioCheckbox.className + " input-accordion-checkbox";
+    span.insertBefore(visibleCheckbox, span.firstChild);
+
+    accordion.visibleCheckbox = visibleCheckbox;
+    accordion.onVisibleCheckboxChange = function() {
+        if (linked && isOpen() != visibleCheckbox.checked) {
+            labelWrap.click();
+        }
+
+        gradioCheckbox.checked = visibleCheckbox.checked;
+        updateInput(gradioCheckbox);
+    };
+
+    visibleCheckbox.addEventListener('click', function(event) {
+        linked = false;
+        event.stopPropagation();
+    });
+    visibleCheckbox.addEventListener('input', accordion.onVisibleCheckboxChange);
 }
 
 onUiLoaded(function() {
     for (var accordion of gradioApp().querySelectorAll('.input-accordion')) {
-        var labelWrap = accordion.querySelector('.label-wrap');
-        observerAccordionOpen.observe(labelWrap, {attributes: true, attributeFilter: ['class']});
-
-        var extra = gradioApp().querySelector('#' + accordion.id + "-extra");
-        if (extra) {
-            labelWrap.insertBefore(extra, labelWrap.lastElementChild);
-        }
+        setupAccordion(accordion);
     }
 });

modules/api/api.py

@@ -17,15 +17,14 @@ from fastapi.encoders import jsonable_encoder
 from secrets import compare_digest
 
 import modules.shared as shared
-from modules import sd_samplers, deepbooru, sd_hijack, images, scripts, ui, postprocessing, errors, restart, shared_items, script_callbacks, generation_parameters_copypaste
+from modules import sd_samplers, deepbooru, sd_hijack, images, scripts, ui, postprocessing, errors, restart, shared_items, script_callbacks, generation_parameters_copypaste, sd_models
 from modules.api import models
 from modules.shared import opts
 from modules.processing import StableDiffusionProcessingTxt2Img, StableDiffusionProcessingImg2Img, process_images
 from modules.textual_inversion.textual_inversion import create_embedding, train_embedding
 from modules.textual_inversion.preprocess import preprocess
 from modules.hypernetworks.hypernetwork import create_hypernetwork, train_hypernetwork
-from PIL import PngImagePlugin,Image
-from modules.sd_models import unload_model_weights, reload_model_weights, checkpoint_aliases
+from PIL import PngImagePlugin, Image
 from modules.sd_models_config import find_checkpoint_config_near_filename
 from modules.realesrgan_model import get_realesrgan_models
 from modules import devices
@@ -103,7 +102,8 @@ def decode_base64_to_image(encoding):
 
 def encode_pil_to_base64(image):
     with io.BytesIO() as output_bytes:
-
+        if isinstance(image, str):
+            return image
         if opts.samples_format.lower() == 'png':
             use_metadata = False
             metadata = PngImagePlugin.PngInfo()
@@ -540,12 +540,12 @@ class Api:
         return {}
 
     def unloadapi(self):
-        unload_model_weights()
+        sd_models.unload_model_weights()
 
         return {}
 
     def reloadapi(self):
-        reload_model_weights()
+        sd_models.send_model_to_device(shared.sd_model)
 
         return {}
 
@@ -565,7 +565,7 @@ class Api:
 
     def set_config(self, req: dict[str, Any]):
         checkpoint_name = req.get("sd_model_checkpoint", None)
-        if checkpoint_name is not None and checkpoint_name not in checkpoint_aliases:
+        if checkpoint_name is not None and checkpoint_name not in sd_models.checkpoint_aliases:
             raise RuntimeError(f"model {checkpoint_name!r} not found")
 
         for k, v in req.items():

modules/cmd_args.py

@@ -107,7 +107,7 @@ parser.add_argument("--tls-certfile", type=str, help="Partially enables TLS, req
 parser.add_argument("--disable-tls-verify", action="store_false", help="When passed, enables the use of self-signed certificates.", default=None)
 parser.add_argument("--server-name", type=str, help="Sets hostname of server", default=None)
 parser.add_argument("--gradio-queue", action='store_true', help="does not do anything", default=True)
-parser.add_argument("--no-gradio-queue", action='store_true', help="Disables gradio queue; causes the webpage to use http requests instead of websockets; was the defaul in earlier versions")
+parser.add_argument("--no-gradio-queue", action='store_true', help="Disables gradio queue; causes the webpage to use http requests instead of websockets; was the default in earlier versions")
 parser.add_argument("--skip-version-check", action='store_true', help="Do not check versions of torch and xformers")
 parser.add_argument("--no-hashing", action='store_true', help="disable sha256 hashing of checkpoints to help loading performance", default=False)
 parser.add_argument("--no-download-sd-model", action='store_true', help="don't download SD1.5 model even if no model is found in --ckpt-dir", default=False)

modules/gfpgan_model.py

@@ -9,6 +9,7 @@ from modules import paths, shared, devices, modelloader, errors
 model_dir = "GFPGAN"
 user_path = None
 model_path = os.path.join(paths.models_path, model_dir)
+model_file_path = None
 model_url = "https://github.com/TencentARC/GFPGAN/releases/download/v1.3.0/GFPGANv1.4.pth"
 have_gfpgan = False
 loaded_gfpgan_model = None
@@ -17,6 +18,7 @@ loaded_gfpgan_model = None
 def gfpgann():
     global loaded_gfpgan_model
     global model_path
+    global model_file_path
     if loaded_gfpgan_model is not None:
         loaded_gfpgan_model.gfpgan.to(devices.device_gfpgan)
         return loaded_gfpgan_model
@@ -24,17 +26,24 @@ def gfpgann():
     if gfpgan_constructor is None:
         return None
 
-    models = modelloader.load_models(model_path, model_url, user_path, ext_filter="GFPGAN")
+    models = modelloader.load_models(model_path, model_url, user_path, ext_filter=['.pth'])
+
     if len(models) == 1 and models[0].startswith("http"):
         model_file = models[0]
     elif len(models) != 0:
-        latest_file = max(models, key=os.path.getctime)
+        gfp_models = []
+        for item in models:
+            if 'GFPGAN' in os.path.basename(item):
+                gfp_models.append(item)
+        latest_file = max(gfp_models, key=os.path.getctime)
         model_file = latest_file
     else:
         print("Unable to load gfpgan model!")
         return None
+
     if hasattr(facexlib.detection.retinaface, 'device'):
         facexlib.detection.retinaface.device = devices.device_gfpgan
+    model_file_path = model_file
     model = gfpgan_constructor(model_path=model_file, upscale=1, arch='clean', channel_multiplier=2, bg_upsampler=None, device=devices.device_gfpgan)
     loaded_gfpgan_model = model
 
@@ -77,19 +86,25 @@ def setup_model(dirname):
         global user_path
         global have_gfpgan
         global gfpgan_constructor
+        global model_file_path
+
+        facexlib_path = model_path
+
+        if dirname is not None:
+            facexlib_path = dirname
 
         load_file_from_url_orig = gfpgan.utils.load_file_from_url
         facex_load_file_from_url_orig = facexlib.detection.load_file_from_url
         facex_load_file_from_url_orig2 = facexlib.parsing.load_file_from_url
 
         def my_load_file_from_url(**kwargs):
-            return load_file_from_url_orig(**dict(kwargs, model_dir=model_path))
+            return load_file_from_url_orig(**dict(kwargs, model_dir=model_file_path))
 
         def facex_load_file_from_url(**kwargs):
-            return facex_load_file_from_url_orig(**dict(kwargs, save_dir=model_path, model_dir=None))
+            return facex_load_file_from_url_orig(**dict(kwargs, save_dir=facexlib_path, model_dir=None))
 
         def facex_load_file_from_url2(**kwargs):
-            return facex_load_file_from_url_orig2(**dict(kwargs, save_dir=model_path, model_dir=None))
+            return facex_load_file_from_url_orig2(**dict(kwargs, save_dir=facexlib_path, model_dir=None))
 
         gfpgan.utils.load_file_from_url = my_load_file_from_url
         facexlib.detection.load_file_from_url = facex_load_file_from_url

modules/initialize_util.py

@@ -150,10 +150,14 @@ def dumpstacks():
 
 def configure_sigint_handler():
     # make the program just exit at ctrl+c without waiting for anything
+
+    from modules import shared
+
     def sigint_handler(sig, frame):
         print(f'Interrupted with signal {sig} in {frame}')
 
-        dumpstacks()
+        if shared.opts.dump_stacks_on_signal:
+            dumpstacks()
 
         os._exit(0)
 

modules/processing.py

@@ -711,7 +711,7 @@ def process_images(p: StableDiffusionProcessing) -> Processed:
     if p.scripts is not None:
         p.scripts.before_process(p)
 
-    stored_opts = {k: opts.data[k] for k in p.override_settings.keys() if k in opts.data}
+    stored_opts = {k: opts.data[k] if k in opts.data else opts.get_default(k) for k in p.override_settings.keys() if k in opts.data}
 
     try:
         # if no checkpoint override or the override checkpoint can't be found, remove override entry and load opts checkpoint
@@ -886,6 +886,8 @@ def process_images_inner(p: StableDiffusionProcessing) -> Processed:
 
             devices.torch_gc()
 
+            state.nextjob()
+
             if p.scripts is not None:
                 p.scripts.postprocess_batch(p, x_samples_ddim, batch_number=n)
 
@@ -958,7 +960,8 @@ def process_images_inner(p: StableDiffusionProcessing) -> Processed:
 
             devices.torch_gc()
 
-            state.nextjob()
+        if not infotexts:
+            infotexts.append(Processed(p, []).infotext(p, 0))
 
         p.color_corrections = None
 

modules/prompt_parser.py

@@ -4,7 +4,7 @@ import re
 from collections import namedtuple
 import lark
 
-# a prompt like this: "fantasy landscape with a [mountain:lake:0.25] and [an oak:a christmas tree:0.75][ in foreground::0.6][ in background:0.25] [shoddy:masterful:0.5]"
+# a prompt like this: "fantasy landscape with a [mountain:lake:0.25] and [an oak:a christmas tree:0.75][ in foreground::0.6][: in background:0.25] [shoddy:masterful:0.5]"
 # will be represented with prompt_schedule like this (assuming steps=100):
 # [25, 'fantasy landscape with a mountain and an oak in foreground shoddy']
 # [50, 'fantasy landscape with a lake and an oak in foreground in background shoddy']

modules/sd_hijack.py

@@ -5,7 +5,7 @@ from types import MethodType
 from modules import devices, sd_hijack_optimizations, shared, script_callbacks, errors, sd_unet, patches
 from modules.hypernetworks import hypernetwork
 from modules.shared import cmd_opts
-from modules import sd_hijack_clip, sd_hijack_open_clip, sd_hijack_unet, sd_hijack_xlmr, xlmr
+from modules import sd_hijack_clip, sd_hijack_open_clip, sd_hijack_unet, sd_hijack_xlmr, xlmr, xlmr_m18
 
 import ldm.modules.attention
 import ldm.modules.diffusionmodules.model
@@ -184,6 +184,20 @@ class StableDiffusionModelHijack:
             errors.display(e, "applying cross attention optimization")
             undo_optimizations()
 
+    def convert_sdxl_to_ssd(self, m):
+        """Converts an SDXL model to a Segmind Stable Diffusion model (see https://huggingface.co/segmind/SSD-1B)"""
+
+        delattr(m.model.diffusion_model.middle_block, '1')
+        delattr(m.model.diffusion_model.middle_block, '2')
+        for i in ['9', '8', '7', '6', '5', '4']:
+            delattr(m.model.diffusion_model.input_blocks[7][1].transformer_blocks, i)
+            delattr(m.model.diffusion_model.input_blocks[8][1].transformer_blocks, i)
+            delattr(m.model.diffusion_model.output_blocks[0][1].transformer_blocks, i)
+            delattr(m.model.diffusion_model.output_blocks[1][1].transformer_blocks, i)
+        delattr(m.model.diffusion_model.output_blocks[4][1].transformer_blocks, '1')
+        delattr(m.model.diffusion_model.output_blocks[5][1].transformer_blocks, '1')
+        devices.torch_gc()
+
     def hijack(self, m):
         conditioner = getattr(m, 'conditioner', None)
         if conditioner:
@@ -211,7 +225,7 @@ class StableDiffusionModelHijack:
             else:
                 m.cond_stage_model = conditioner
 
-        if type(m.cond_stage_model) == xlmr.BertSeriesModelWithTransformation:
+        if type(m.cond_stage_model) == xlmr.BertSeriesModelWithTransformation or type(m.cond_stage_model) == xlmr_m18.BertSeriesModelWithTransformation:
             model_embeddings = m.cond_stage_model.roberta.embeddings
             model_embeddings.token_embedding = EmbeddingsWithFixes(model_embeddings.word_embeddings, self)
             m.cond_stage_model = sd_hijack_xlmr.FrozenXLMREmbedderWithCustomWords(m.cond_stage_model, self)
@@ -242,8 +256,12 @@ class StableDiffusionModelHijack:
 
         self.layers = flatten(m)
 
+        import modules.models.diffusion.ddpm_edit
+
         if isinstance(m, ldm.models.diffusion.ddpm.LatentDiffusion):
             sd_unet.original_forward = ldm_original_forward
+        elif isinstance(m, modules.models.diffusion.ddpm_edit.LatentDiffusion):
+            sd_unet.original_forward = ldm_original_forward
         elif isinstance(m, sgm.models.diffusion.DiffusionEngine):
             sd_unet.original_forward = sgm_original_forward
         else:

modules/sd_models.py

@@ -1,7 +1,6 @@
 import collections
 import os.path
 import sys
-import gc
 import threading
 
 import torch
@@ -353,16 +352,19 @@ def load_model_weights(model, checkpoint_info: CheckpointInfo, state_dict, timer
     model.is_sdxl = hasattr(model, 'conditioner')
     model.is_sd2 = not model.is_sdxl and hasattr(model.cond_stage_model, 'model')
     model.is_sd1 = not model.is_sdxl and not model.is_sd2
-
+    model.is_ssd = model.is_sdxl and 'model.diffusion_model.middle_block.1.transformer_blocks.0.attn1.to_q.weight' not in state_dict.keys()
     if model.is_sdxl:
         sd_models_xl.extend_sdxl(model)
 
-    model.load_state_dict(state_dict, strict=False)
-    timer.record("apply weights to model")
+    if model.is_ssd:
+        sd_hijack.model_hijack.convert_sdxl_to_ssd(model)
 
     if shared.opts.sd_checkpoint_cache > 0:
         # cache newly loaded model
-        checkpoints_loaded[checkpoint_info] = state_dict
+        checkpoints_loaded[checkpoint_info] = state_dict.copy()
+
+    model.load_state_dict(state_dict, strict=False)
+    timer.record("apply weights to model")
 
     del state_dict
 
@@ -798,17 +800,7 @@ def reload_model_weights(sd_model=None, info=None):
 
 
 def unload_model_weights(sd_model=None, info=None):
-    timer = Timer()
-
-    if model_data.sd_model:
-        model_data.sd_model.to(devices.cpu)
-        sd_hijack.model_hijack.undo_hijack(model_data.sd_model)
-        model_data.sd_model = None
-        sd_model = None
-        gc.collect()
-        devices.torch_gc()
-
-    print(f"Unloaded weights {timer.summary()}.")
+    send_model_to_cpu(sd_model or shared.sd_model)
 
     return sd_model
 

modules/sd_models_config.py

@@ -21,7 +21,7 @@ config_unopenclip = os.path.join(sd_repo_configs_path, "v2-1-stable-unclip-h-inf
 config_inpainting = os.path.join(sd_configs_path, "v1-inpainting-inference.yaml")
 config_instruct_pix2pix = os.path.join(sd_configs_path, "instruct-pix2pix.yaml")
 config_alt_diffusion = os.path.join(sd_configs_path, "alt-diffusion-inference.yaml")
-
+config_alt_diffusion_m18 = os.path.join(sd_configs_path, "alt-diffusion-m18-inference.yaml")
 
 def is_using_v_parameterization_for_sd2(state_dict):
     """
@@ -95,7 +95,10 @@ def guess_model_config_from_state_dict(sd, filename):
         if diffusion_model_input.shape[1] == 8:
             return config_instruct_pix2pix
 
+
     if sd.get('cond_stage_model.roberta.embeddings.word_embeddings.weight', None) is not None:
+        if sd.get('cond_stage_model.transformation.weight').size()[0] == 1024:
+            return config_alt_diffusion_m18
         return config_alt_diffusion
 
     return config_default

modules/sd_models_types.py

@@ -22,7 +22,10 @@ class WebuiSdModel(LatentDiffusion):
     """structure with additional information about the file with model's weights"""
 
     is_sdxl: bool
-    """True if the model's architecture is SDXL"""
+    """True if the model's architecture is SDXL or SSD"""
+
+    is_ssd: bool
+    """True if the model is SSD"""
 
     is_sd2: bool
     """True if the model's architecture is SD 2.x"""

modules/sd_samplers_common.py

@@ -3,7 +3,7 @@ from collections import namedtuple
 import numpy as np
 import torch
 from PIL import Image
-from modules import devices, images, sd_vae_approx, sd_samplers, sd_vae_taesd, shared, sd_models
+from modules import devices, images, sd_vae_approx, sd_samplers, sd_vae_taesd, sd_vae_consistency, shared, sd_models
 from modules.shared import opts, state
 import k_diffusion.sampling
 
@@ -31,7 +31,7 @@ def setup_img2img_steps(p, steps=None):
     return steps, t_enc
 
 
-approximation_indexes = {"Full": 0, "Approx NN": 1, "Approx cheap": 2, "TAESD": 3}
+approximation_indexes = {"Full": 0, "Approx NN": 1, "Approx cheap": 2, "TAESD": 3, "Consistency Decoder": 4}
 
 
 def samples_to_images_tensor(sample, approximation=None, model=None):
@@ -51,6 +51,13 @@ def samples_to_images_tensor(sample, approximation=None, model=None):
     elif approximation == 3:
         x_sample = sd_vae_taesd.decoder_model()(sample.to(devices.device, devices.dtype)).detach()
         x_sample = x_sample * 2 - 1
+    elif approximation == 4:
+        with devices.autocast(), torch.no_grad():
+            x_sample = sd_vae_consistency.decoder_model()(
+                sample.detach().to(devices.device, devices.dtype)/0.18215,
+                schedule=[float(i.strip()) for i in shared.opts.sd_vae_consistency_schedule.split(',')],
+            )
+        sd_vae_consistency.unload()
     else:
         if model is None:
             model = shared.sd_model

modules/sd_vae_consistency.py

@@ -0,0 +1,34 @@
+"""
+Consistency Decoder
+Improved decoding for stable diffusion vaes.
+
+https://github.com/openai/consistencydecoder
+"""
+import os
+
+from modules import devices, paths_internal, shared
+from consistencydecoder import ConsistencyDecoder
+
+
+sd_vae_consistency_models = None
+model_path = os.path.join(paths_internal.models_path, 'consistencydecoder')
+
+
+def decoder_model():
+    global sd_vae_consistency_models
+    if getattr(shared.sd_model, 'is_sdxl', False):
+        raise NotImplementedError("SDXL is not supported for consistency decoder")
+    if sd_vae_consistency_models is not None:
+        sd_vae_consistency_models.ckpt.to(devices.device)
+        return sd_vae_consistency_models
+
+    loaded_model = ConsistencyDecoder(devices.device, model_path)
+    sd_vae_consistency_models = loaded_model
+    return loaded_model
+
+
+def unload():
+    global sd_vae_consistency_models
+    if sd_vae_consistency_models is not None:
+        devices.torch_gc()
+        sd_vae_consistency_models.ckpt.to('cpu')

modules/shared_items.py

@@ -67,6 +67,8 @@ def reload_hypernetworks():
 
 
 ui_reorder_categories_builtin_items = [
+    "prompt",
+    "image",
     "inpaint",
     "sampler",
     "accordions",

modules/shared_options.py

@@ -62,6 +62,8 @@ options_templates.update(options_section(('saving-images', "Saving images/grids"
     "clean_temp_dir_at_start": OptionInfo(False, "Cleanup non-default temporary directory when starting webui"),
 
     "save_incomplete_images": OptionInfo(False, "Save incomplete images").info("save images that has been interrupted in mid-generation; even if not saved, they will still show up in webui output."),
+
+    "notification_audio": OptionInfo(True, "Play notification sound after image generation").info("notification.mp3 should be present in the root directory").needs_reload_ui(),
 }))
 
 options_templates.update(options_section(('saving-paths', "Paths for saving"), {
@@ -110,6 +112,7 @@ options_templates.update(options_section(('system', "System"), {
     "list_hidden_files": OptionInfo(True, "Load models/files in hidden directories").info("directory is hidden if its name starts with \".\""),
     "disable_mmap_load_safetensors": OptionInfo(False, "Disable memmapping for loading .safetensors files.").info("fixes very slow loading speed in some cases"),
     "hide_ldm_prints": OptionInfo(True, "Prevent Stability-AI's ldm/sgm modules from printing noise to console."),
+    "dump_stacks_on_signal": OptionInfo(False, "Print stack traces before exiting the program with ctrl+c."),
 }))
 
 options_templates.update(options_section(('API', "API"), {
@@ -169,7 +172,8 @@ For img2img, VAE is used to process user's input image before the sampling, and
     "sd_vae_overrides_per_model_preferences": OptionInfo(True, "Selected VAE overrides per-model preferences").info("you can set per-model VAE either by editing user metadata for checkpoints, or by making the VAE have same name as checkpoint"),
     "auto_vae_precision": OptionInfo(True, "Automatically revert VAE to 32-bit floats").info("triggers when a tensor with NaNs is produced in VAE; disabling the option in this case will result in a black square image"),
     "sd_vae_encode_method": OptionInfo("Full", "VAE type for encode", gr.Radio, {"choices": ["Full", "TAESD"]}, infotext='VAE Encoder').info("method to encode image to latent (use in img2img, hires-fix or inpaint mask)"),
-    "sd_vae_decode_method": OptionInfo("Full", "VAE type for decode", gr.Radio, {"choices": ["Full", "TAESD"]}, infotext='VAE Decoder').info("method to decode latent to image"),
+    "sd_vae_decode_method": OptionInfo("Full", "VAE type for decode", gr.Radio, {"choices": ["Full", "TAESD", "Consistency Decoder"]}, infotext='VAE Decoder').info("method to decode latent to image"),
+    "sd_vae_consistency_schedule": OptionInfo("1.0, 0.5", "consistency schedule").info("sampling schedule for consistency decoder."),
 }))
 
 options_templates.update(options_section(('img2img', "img2img"), {
@@ -231,6 +235,8 @@ options_templates.update(options_section(('extra_networks', "Extra Networks"), {
     "extra_networks_card_height": OptionInfo(0, "Card height for Extra Networks").info("in pixels"),
     "extra_networks_card_text_scale": OptionInfo(1.0, "Card text scale", gr.Slider, {"minimum": 0.0, "maximum": 2.0, "step": 0.01}).info("1 = original size"),
     "extra_networks_card_show_desc": OptionInfo(True, "Show description on card"),
+    "extra_networks_card_order_field": OptionInfo("Name", "Default order field for Extra Networks cards", gr.Dropdown, {"choices": ['Name', 'Date Created', 'Date Modified']}).needs_reload_ui(),
+    "extra_networks_card_order": OptionInfo("Ascending", "Default order for Extra Networks cards", gr.Dropdown, {"choices": ['Ascending', 'Descending']}).needs_reload_ui(),
     "extra_networks_add_text_separator": OptionInfo(" ", "Extra networks separator").info("extra text to add before <...> when adding extra network to prompt"),
     "ui_extra_networks_tab_reorder": OptionInfo("", "Extra networks tab order").needs_reload_ui(),
     "textual_inversion_print_at_load": OptionInfo(False, "Print a list of Textual Inversion embeddings when loading model"),
@@ -267,6 +273,7 @@ options_templates.update(options_section(('ui', "User interface"), {
     "hires_fix_show_sampler": OptionInfo(False, "Hires fix: show hires checkpoint and sampler selection").needs_reload_ui(),
     "hires_fix_show_prompts": OptionInfo(False, "Hires fix: show hires prompt and negative prompt").needs_reload_ui(),
     "disable_token_counters": OptionInfo(False, "Disable prompt token counters").needs_reload_ui(),
+    "compact_prompt_box": OptionInfo(False, "Compact prompt layout").info("puts prompt and negative prompt inside the Generate tab, leaving more vertical space for the image on the right").needs_reload_ui(),
 }))
 
 
@@ -332,4 +339,3 @@ options_templates.update(options_section((None, "Hidden options"), {
     "restore_config_state_file": OptionInfo("", "Config state file to restore from, under 'config-states/' folder"),
     "sd_checkpoint_hash": OptionInfo("", "SHA256 hash of the current checkpoint"),
 }))
-

modules/textual_inversion/textual_inversion.py

@@ -181,40 +181,7 @@ class EmbeddingDatabase:
         else:
             return
 
-
-        # textual inversion embeddings
-        if 'string_to_param' in data:
-            param_dict = data['string_to_param']
-            param_dict = getattr(param_dict, '_parameters', param_dict)  # fix for torch 1.12.1 loading saved file from torch 1.11
-            assert len(param_dict) == 1, 'embedding file has multiple terms in it'
-            emb = next(iter(param_dict.items()))[1]
-            vec = emb.detach().to(devices.device, dtype=torch.float32)
-            shape = vec.shape[-1]
-            vectors = vec.shape[0]
-        elif type(data) == dict and 'clip_g' in data and 'clip_l' in data:  # SDXL embedding
-            vec = {k: v.detach().to(devices.device, dtype=torch.float32) for k, v in data.items()}
-            shape = data['clip_g'].shape[-1] + data['clip_l'].shape[-1]
-            vectors = data['clip_g'].shape[0]
-        elif type(data) == dict and type(next(iter(data.values()))) == torch.Tensor: # diffuser concepts
-            assert len(data.keys()) == 1, 'embedding file has multiple terms in it'
-
-            emb = next(iter(data.values()))
-            if len(emb.shape) == 1:
-                emb = emb.unsqueeze(0)
-            vec = emb.detach().to(devices.device, dtype=torch.float32)
-            shape = vec.shape[-1]
-            vectors = vec.shape[0]
-        else:
-            raise Exception(f"Couldn't identify {filename} as neither textual inversion embedding nor diffuser concept.")
-
-        embedding = Embedding(vec, name)
-        embedding.step = data.get('step', None)
-        embedding.sd_checkpoint = data.get('sd_checkpoint', None)
-        embedding.sd_checkpoint_name = data.get('sd_checkpoint_name', None)
-        embedding.vectors = vectors
-        embedding.shape = shape
-        embedding.filename = path
-        embedding.set_hash(hashes.sha256(embedding.filename, "textual_inversion/" + name) or '')
+        embedding = create_embedding_from_data(data, name, filename=filename, filepath=path)
 
         if self.expected_shape == -1 or self.expected_shape == embedding.shape:
             self.register_embedding(embedding, shared.sd_model)
@@ -313,6 +280,45 @@ def create_embedding(name, num_vectors_per_token, overwrite_old, init_text='*'):
     return fn
 
 
+def create_embedding_from_data(data, name, filename='unknown embedding file', filepath=None):
+    if 'string_to_param' in data:  # textual inversion embeddings
+        param_dict = data['string_to_param']
+        param_dict = getattr(param_dict, '_parameters', param_dict)  # fix for torch 1.12.1 loading saved file from torch 1.11
+        assert len(param_dict) == 1, 'embedding file has multiple terms in it'
+        emb = next(iter(param_dict.items()))[1]
+        vec = emb.detach().to(devices.device, dtype=torch.float32)
+        shape = vec.shape[-1]
+        vectors = vec.shape[0]
+    elif type(data) == dict and 'clip_g' in data and 'clip_l' in data:  # SDXL embedding
+        vec = {k: v.detach().to(devices.device, dtype=torch.float32) for k, v in data.items()}
+        shape = data['clip_g'].shape[-1] + data['clip_l'].shape[-1]
+        vectors = data['clip_g'].shape[0]
+    elif type(data) == dict and type(next(iter(data.values()))) == torch.Tensor:  # diffuser concepts
+        assert len(data.keys()) == 1, 'embedding file has multiple terms in it'
+
+        emb = next(iter(data.values()))
+        if len(emb.shape) == 1:
+            emb = emb.unsqueeze(0)
+        vec = emb.detach().to(devices.device, dtype=torch.float32)
+        shape = vec.shape[-1]
+        vectors = vec.shape[0]
+    else:
+        raise Exception(f"Couldn't identify {filename} as neither textual inversion embedding nor diffuser concept.")
+
+    embedding = Embedding(vec, name)
+    embedding.step = data.get('step', None)
+    embedding.sd_checkpoint = data.get('sd_checkpoint', None)
+    embedding.sd_checkpoint_name = data.get('sd_checkpoint_name', None)
+    embedding.vectors = vectors
+    embedding.shape = shape
+
+    if filepath:
+        embedding.filename = filepath
+        embedding.set_hash(hashes.sha256(filepath, "textual_inversion/" + name) or '')
+
+    return embedding
+
+
 def write_loss(log_directory, filename, step, epoch_len, values):
     if shared.opts.training_write_csv_every == 0:
         return

modules/ui.py

@@ -12,7 +12,7 @@ from PIL import Image, PngImagePlugin  # noqa: F401
 from modules.call_queue import wrap_gradio_gpu_call, wrap_queued_call, wrap_gradio_call
 
 from modules import gradio_extensons  # noqa: F401
-from modules import sd_hijack, sd_models, script_callbacks, ui_extensions, deepbooru, extra_networks, ui_common, ui_postprocessing, progress, ui_loadsave, shared_items, ui_settings, timer, sysinfo, ui_checkpoint_merger, ui_prompt_styles, scripts, sd_samplers, processing, ui_extra_networks
+from modules import sd_hijack, sd_models, script_callbacks, ui_extensions, deepbooru, extra_networks, ui_common, ui_postprocessing, progress, ui_loadsave, shared_items, ui_settings, timer, sysinfo, ui_checkpoint_merger, scripts, sd_samplers, processing, ui_extra_networks, ui_toprow
 from modules.ui_components import FormRow, FormGroup, ToolButton, FormHTML, InputAccordion, ResizeHandleRow
 from modules.paths import script_path
 from modules.ui_common import create_refresh_button
@@ -25,7 +25,6 @@ import modules.hypernetworks.ui as hypernetworks_ui
 import modules.textual_inversion.ui as textual_inversion_ui
 import modules.textual_inversion.textual_inversion as textual_inversion
 import modules.shared as shared
-import modules.images
 from modules import prompt_parser
 from modules.sd_hijack import model_hijack
 from modules.generation_parameters_copypaste import image_from_url_text
@@ -177,79 +176,6 @@ def update_negative_prompt_token_counter(text, steps):
     return update_token_counter(text, steps, is_positive=False)
 
 
-class Toprow:
-    """Creates a top row UI with prompts, generate button, styles, extra little buttons for things, and enables some functionality related to their operation"""
-
-    def __init__(self, is_img2img):
-        id_part = "img2img" if is_img2img else "txt2img"
-        self.id_part = id_part
-
-        with gr.Row(elem_id=f"{id_part}_toprow", variant="compact"):
-            with gr.Column(elem_id=f"{id_part}_prompt_container", scale=6):
-                with gr.Row():
-                    with gr.Column(scale=80):
-                        with gr.Row():
-                            self.prompt = gr.Textbox(label="Prompt", elem_id=f"{id_part}_prompt", show_label=False, lines=3, placeholder="Prompt (press Ctrl+Enter or Alt+Enter to generate)", elem_classes=["prompt"])
-                            self.prompt_img = gr.File(label="", elem_id=f"{id_part}_prompt_image", file_count="single", type="binary", visible=False)
-
-                with gr.Row():
-                    with gr.Column(scale=80):
-                        with gr.Row():
-                            self.negative_prompt = gr.Textbox(label="Negative prompt", elem_id=f"{id_part}_neg_prompt", show_label=False, lines=3, placeholder="Negative prompt (press Ctrl+Enter or Alt+Enter to generate)", elem_classes=["prompt"])
-
-            self.button_interrogate = None
-            self.button_deepbooru = None
-            if is_img2img:
-                with gr.Column(scale=1, elem_classes="interrogate-col"):
-                    self.button_interrogate = gr.Button('Interrogate\nCLIP', elem_id="interrogate")
-                    self.button_deepbooru = gr.Button('Interrogate\nDeepBooru', elem_id="deepbooru")
-
-            with gr.Column(scale=1, elem_id=f"{id_part}_actions_column"):
-                with gr.Row(elem_id=f"{id_part}_generate_box", elem_classes="generate-box"):
-                    self.interrupt = gr.Button('Interrupt', elem_id=f"{id_part}_interrupt", elem_classes="generate-box-interrupt")
-                    self.skip = gr.Button('Skip', elem_id=f"{id_part}_skip", elem_classes="generate-box-skip")
-                    self.submit = gr.Button('Generate', elem_id=f"{id_part}_generate", variant='primary')
-
-                    self.skip.click(
-                        fn=lambda: shared.state.skip(),
-                        inputs=[],
-                        outputs=[],
-                    )
-
-                    self.interrupt.click(
-                        fn=lambda: shared.state.interrupt(),
-                        inputs=[],
-                        outputs=[],
-                    )
-
-                with gr.Row(elem_id=f"{id_part}_tools"):
-                    self.paste = ToolButton(value=paste_symbol, elem_id="paste", tooltip="Read generation parameters from prompt or last generation if prompt is empty into user interface.")
-                    self.clear_prompt_button = ToolButton(value=clear_prompt_symbol, elem_id=f"{id_part}_clear_prompt", tooltip="Clear prompt")
-                    self.apply_styles = ToolButton(value=ui_prompt_styles.styles_materialize_symbol, elem_id=f"{id_part}_style_apply", tooltip="Apply all selected styles to prompts.")
-                    self.restore_progress_button = ToolButton(value=restore_progress_symbol, elem_id=f"{id_part}_restore_progress", visible=False, tooltip="Restore progress")
-
-                    self.token_counter = gr.HTML(value="<span>0/75</span>", elem_id=f"{id_part}_token_counter", elem_classes=["token-counter"])
-                    self.token_button = gr.Button(visible=False, elem_id=f"{id_part}_token_button")
-                    self.negative_token_counter = gr.HTML(value="<span>0/75</span>", elem_id=f"{id_part}_negative_token_counter", elem_classes=["token-counter"])
-                    self.negative_token_button = gr.Button(visible=False, elem_id=f"{id_part}_negative_token_button")
-
-                    self.clear_prompt_button.click(
-                        fn=lambda *x: x,
-                        _js="confirm_clear_prompt",
-                        inputs=[self.prompt, self.negative_prompt],
-                        outputs=[self.prompt, self.negative_prompt],
-                    )
-
-                self.ui_styles = ui_prompt_styles.UiPromptStyles(id_part, self.prompt, self.negative_prompt)
-                self.ui_styles.setup_apply_button(self.apply_styles)
-
-        self.prompt_img.change(
-            fn=modules.images.image_data,
-            inputs=[self.prompt_img],
-            outputs=[self.prompt, self.prompt_img],
-            show_progress=False,
-        )
-
 
 def setup_progressbar(*args, **kwargs):
     pass
@@ -288,8 +214,8 @@ def apply_setting(key, value):
     return getattr(opts, key)
 
 
-def create_output_panel(tabname, outdir):
-    return ui_common.create_output_panel(tabname, outdir)
+def create_output_panel(tabname, outdir, toprow=None):
+    return ui_common.create_output_panel(tabname, outdir, toprow)
 
 
 def create_sampler_and_steps_selection(choices, tabname):
@@ -336,7 +262,7 @@ def create_ui():
     scripts.scripts_txt2img.initialize_scripts(is_img2img=False)
 
     with gr.Blocks(analytics_enabled=False) as txt2img_interface:
-        toprow = Toprow(is_img2img=False)
+        toprow = ui_toprow.Toprow(is_img2img=False, is_compact=shared.opts.compact_prompt_box)
 
         dummy_component = gr.Label(visible=False)
 
@@ -348,6 +274,9 @@ def create_ui():
                 scripts.scripts_txt2img.prepare_ui()
 
                 for category in ordered_ui_categories():
+                    if category == "prompt":
+                        toprow.create_inline_toprow_prompts()
+
                     if category == "sampler":
                         steps, sampler_name = create_sampler_and_steps_selection(sd_samplers.visible_sampler_names(), "txt2img")
 
@@ -442,7 +371,7 @@ def create_ui():
                     show_progress=False,
                 )
 
-            txt2img_gallery, generation_info, html_info, html_log = create_output_panel("txt2img", opts.outdir_txt2img_samples)
+            txt2img_gallery, generation_info, html_info, html_log = create_output_panel("txt2img", opts.outdir_txt2img_samples, toprow)
 
             txt2img_args = dict(
                 fn=wrap_gradio_gpu_call(modules.txt2img.txt2img, extra_outputs=[None, '', '']),
@@ -554,7 +483,7 @@ def create_ui():
     scripts.scripts_img2img.initialize_scripts(is_img2img=True)
 
     with gr.Blocks(analytics_enabled=False) as img2img_interface:
-        toprow = Toprow(is_img2img=True)
+        toprow = ui_toprow.Toprow(is_img2img=True, is_compact=shared.opts.compact_prompt_box)
 
         extra_tabs = gr.Tabs(elem_id="img2img_extra_tabs")
         extra_tabs.__enter__()
@@ -577,85 +506,89 @@ def create_ui():
                             button = gr.Button(title)
                             copy_image_buttons.append((button, name, elem))
 
-                with gr.Tabs(elem_id="mode_img2img"):
-                    img2img_selected_tab = gr.State(0)
-
-                    with gr.TabItem('img2img', id='img2img', elem_id="img2img_img2img_tab") as tab_img2img:
-                        init_img = gr.Image(label="Image for img2img", elem_id="img2img_image", show_label=False, source="upload", interactive=True, type="pil", tool="editor", image_mode="RGBA", height=opts.img2img_editor_height)
-                        add_copy_image_controls('img2img', init_img)
-
-                    with gr.TabItem('Sketch', id='img2img_sketch', elem_id="img2img_img2img_sketch_tab") as tab_sketch:
-                        sketch = gr.Image(label="Image for img2img", elem_id="img2img_sketch", show_label=False, source="upload", interactive=True, type="pil", tool="color-sketch", image_mode="RGB", height=opts.img2img_editor_height, brush_color=opts.img2img_sketch_default_brush_color)
-                        add_copy_image_controls('sketch', sketch)
-
-                    with gr.TabItem('Inpaint', id='inpaint', elem_id="img2img_inpaint_tab") as tab_inpaint:
-                        init_img_with_mask = gr.Image(label="Image for inpainting with mask", show_label=False, elem_id="img2maskimg", source="upload", interactive=True, type="pil", tool="sketch", image_mode="RGBA", height=opts.img2img_editor_height, brush_color=opts.img2img_inpaint_mask_brush_color)
-                        add_copy_image_controls('inpaint', init_img_with_mask)
-
-                    with gr.TabItem('Inpaint sketch', id='inpaint_sketch', elem_id="img2img_inpaint_sketch_tab") as tab_inpaint_color:
-                        inpaint_color_sketch = gr.Image(label="Color sketch inpainting", show_label=False, elem_id="inpaint_sketch", source="upload", interactive=True, type="pil", tool="color-sketch", image_mode="RGB", height=opts.img2img_editor_height, brush_color=opts.img2img_inpaint_sketch_default_brush_color)
-                        inpaint_color_sketch_orig = gr.State(None)
-                        add_copy_image_controls('inpaint_sketch', inpaint_color_sketch)
-
-                        def update_orig(image, state):
-                            if image is not None:
-                                same_size = state is not None and state.size == image.size
-                                has_exact_match = np.any(np.all(np.array(image) == np.array(state), axis=-1))
-                                edited = same_size and has_exact_match
-                                return image if not edited or state is None else state
-
-                        inpaint_color_sketch.change(update_orig, [inpaint_color_sketch, inpaint_color_sketch_orig], inpaint_color_sketch_orig)
-
-                    with gr.TabItem('Inpaint upload', id='inpaint_upload', elem_id="img2img_inpaint_upload_tab") as tab_inpaint_upload:
-                        init_img_inpaint = gr.Image(label="Image for img2img", show_label=False, source="upload", interactive=True, type="pil", elem_id="img_inpaint_base")
-                        init_mask_inpaint = gr.Image(label="Mask", source="upload", interactive=True, type="pil", image_mode="RGBA", elem_id="img_inpaint_mask")
-
-                    with gr.TabItem('Batch', id='batch', elem_id="img2img_batch_tab") as tab_batch:
-                        hidden = '<br>Disabled when launched with --hide-ui-dir-config.' if shared.cmd_opts.hide_ui_dir_config else ''
-                        gr.HTML(
-                            "<p style='padding-bottom: 1em;' class=\"text-gray-500\">Process images in a directory on the same machine where the server is running." +
-                            "<br>Use an empty output directory to save pictures normally instead of writing to the output directory." +
-                            f"<br>Add inpaint batch mask directory to enable inpaint batch processing."
-                            f"{hidden}</p>"
-                        )
-                        img2img_batch_input_dir = gr.Textbox(label="Input directory", **shared.hide_dirs, elem_id="img2img_batch_input_dir")
-                        img2img_batch_output_dir = gr.Textbox(label="Output directory", **shared.hide_dirs, elem_id="img2img_batch_output_dir")
-                        img2img_batch_inpaint_mask_dir = gr.Textbox(label="Inpaint batch mask directory (required for inpaint batch processing only)", **shared.hide_dirs, elem_id="img2img_batch_inpaint_mask_dir")
-                        with gr.Accordion("PNG info", open=False):
-                            img2img_batch_use_png_info = gr.Checkbox(label="Append png info to prompts", **shared.hide_dirs, elem_id="img2img_batch_use_png_info")
-                            img2img_batch_png_info_dir = gr.Textbox(label="PNG info directory", **shared.hide_dirs, placeholder="Leave empty to use input directory", elem_id="img2img_batch_png_info_dir")
-                            img2img_batch_png_info_props = gr.CheckboxGroup(["Prompt", "Negative prompt", "Seed", "CFG scale", "Sampler", "Steps", "Model hash"], label="Parameters to take from png info", info="Prompts from png info will be appended to prompts set in ui.")
-
-                    img2img_tabs = [tab_img2img, tab_sketch, tab_inpaint, tab_inpaint_color, tab_inpaint_upload, tab_batch]
-
-                    for i, tab in enumerate(img2img_tabs):
-                        tab.select(fn=lambda tabnum=i: tabnum, inputs=[], outputs=[img2img_selected_tab])
-
-                def copy_image(img):
-                    if isinstance(img, dict) and 'image' in img:
-                        return img['image']
-
-                    return img
-
-                for button, name, elem in copy_image_buttons:
-                    button.click(
-                        fn=copy_image,
-                        inputs=[elem],
-                        outputs=[copy_image_destinations[name]],
-                    )
-                    button.click(
-                        fn=lambda: None,
-                        _js=f"switch_to_{name.replace(' ', '_')}",
-                        inputs=[],
-                        outputs=[],
-                    )
-
-                with FormRow():
-                    resize_mode = gr.Radio(label="Resize mode", elem_id="resize_mode", choices=["Just resize", "Crop and resize", "Resize and fill", "Just resize (latent upscale)"], type="index", value="Just resize")
-
                 scripts.scripts_img2img.prepare_ui()
 
                 for category in ordered_ui_categories():
+                    if category == "prompt":
+                        toprow.create_inline_toprow_prompts()
+
+                    if category == "image":
+                        with gr.Tabs(elem_id="mode_img2img"):
+                            img2img_selected_tab = gr.State(0)
+
+                            with gr.TabItem('img2img', id='img2img', elem_id="img2img_img2img_tab") as tab_img2img:
+                                init_img = gr.Image(label="Image for img2img", elem_id="img2img_image", show_label=False, source="upload", interactive=True, type="pil", tool="editor", image_mode="RGBA", height=opts.img2img_editor_height)
+                                add_copy_image_controls('img2img', init_img)
+
+                            with gr.TabItem('Sketch', id='img2img_sketch', elem_id="img2img_img2img_sketch_tab") as tab_sketch:
+                                sketch = gr.Image(label="Image for img2img", elem_id="img2img_sketch", show_label=False, source="upload", interactive=True, type="pil", tool="color-sketch", image_mode="RGB", height=opts.img2img_editor_height, brush_color=opts.img2img_sketch_default_brush_color)
+                                add_copy_image_controls('sketch', sketch)
+
+                            with gr.TabItem('Inpaint', id='inpaint', elem_id="img2img_inpaint_tab") as tab_inpaint:
+                                init_img_with_mask = gr.Image(label="Image for inpainting with mask", show_label=False, elem_id="img2maskimg", source="upload", interactive=True, type="pil", tool="sketch", image_mode="RGBA", height=opts.img2img_editor_height, brush_color=opts.img2img_inpaint_mask_brush_color)
+                                add_copy_image_controls('inpaint', init_img_with_mask)
+
+                            with gr.TabItem('Inpaint sketch', id='inpaint_sketch', elem_id="img2img_inpaint_sketch_tab") as tab_inpaint_color:
+                                inpaint_color_sketch = gr.Image(label="Color sketch inpainting", show_label=False, elem_id="inpaint_sketch", source="upload", interactive=True, type="pil", tool="color-sketch", image_mode="RGB", height=opts.img2img_editor_height, brush_color=opts.img2img_inpaint_sketch_default_brush_color)
+                                inpaint_color_sketch_orig = gr.State(None)
+                                add_copy_image_controls('inpaint_sketch', inpaint_color_sketch)
+
+                                def update_orig(image, state):
+                                    if image is not None:
+                                        same_size = state is not None and state.size == image.size
+                                        has_exact_match = np.any(np.all(np.array(image) == np.array(state), axis=-1))
+                                        edited = same_size and has_exact_match
+                                        return image if not edited or state is None else state
+
+                                inpaint_color_sketch.change(update_orig, [inpaint_color_sketch, inpaint_color_sketch_orig], inpaint_color_sketch_orig)
+
+                            with gr.TabItem('Inpaint upload', id='inpaint_upload', elem_id="img2img_inpaint_upload_tab") as tab_inpaint_upload:
+                                init_img_inpaint = gr.Image(label="Image for img2img", show_label=False, source="upload", interactive=True, type="pil", elem_id="img_inpaint_base")
+                                init_mask_inpaint = gr.Image(label="Mask", source="upload", interactive=True, type="pil", image_mode="RGBA", elem_id="img_inpaint_mask")
+
+                            with gr.TabItem('Batch', id='batch', elem_id="img2img_batch_tab") as tab_batch:
+                                hidden = '<br>Disabled when launched with --hide-ui-dir-config.' if shared.cmd_opts.hide_ui_dir_config else ''
+                                gr.HTML(
+                                    "<p style='padding-bottom: 1em;' class=\"text-gray-500\">Process images in a directory on the same machine where the server is running." +
+                                    "<br>Use an empty output directory to save pictures normally instead of writing to the output directory." +
+                                    f"<br>Add inpaint batch mask directory to enable inpaint batch processing."
+                                    f"{hidden}</p>"
+                                )
+                                img2img_batch_input_dir = gr.Textbox(label="Input directory", **shared.hide_dirs, elem_id="img2img_batch_input_dir")
+                                img2img_batch_output_dir = gr.Textbox(label="Output directory", **shared.hide_dirs, elem_id="img2img_batch_output_dir")
+                                img2img_batch_inpaint_mask_dir = gr.Textbox(label="Inpaint batch mask directory (required for inpaint batch processing only)", **shared.hide_dirs, elem_id="img2img_batch_inpaint_mask_dir")
+                                with gr.Accordion("PNG info", open=False):
+                                    img2img_batch_use_png_info = gr.Checkbox(label="Append png info to prompts", **shared.hide_dirs, elem_id="img2img_batch_use_png_info")
+                                    img2img_batch_png_info_dir = gr.Textbox(label="PNG info directory", **shared.hide_dirs, placeholder="Leave empty to use input directory", elem_id="img2img_batch_png_info_dir")
+                                    img2img_batch_png_info_props = gr.CheckboxGroup(["Prompt", "Negative prompt", "Seed", "CFG scale", "Sampler", "Steps", "Model hash"], label="Parameters to take from png info", info="Prompts from png info will be appended to prompts set in ui.")
+
+                            img2img_tabs = [tab_img2img, tab_sketch, tab_inpaint, tab_inpaint_color, tab_inpaint_upload, tab_batch]
+
+                            for i, tab in enumerate(img2img_tabs):
+                                tab.select(fn=lambda tabnum=i: tabnum, inputs=[], outputs=[img2img_selected_tab])
+
+                        def copy_image(img):
+                            if isinstance(img, dict) and 'image' in img:
+                                return img['image']
+
+                            return img
+
+                        for button, name, elem in copy_image_buttons:
+                            button.click(
+                                fn=copy_image,
+                                inputs=[elem],
+                                outputs=[copy_image_destinations[name]],
+                            )
+                            button.click(
+                                fn=lambda: None,
+                                _js=f"switch_to_{name.replace(' ', '_')}",
+                                inputs=[],
+                                outputs=[],
+                            )
+
+                        with FormRow():
+                            resize_mode = gr.Radio(label="Resize mode", elem_id="resize_mode", choices=["Just resize", "Crop and resize", "Resize and fill", "Just resize (latent upscale)"], type="index", value="Just resize")
+
                     if category == "sampler":
                         steps, sampler_name = create_sampler_and_steps_selection(sd_samplers.visible_sampler_names(), "img2img")
 
@@ -756,20 +689,20 @@ def create_ui():
                                 with gr.Column(scale=4):
                                     inpaint_full_res_padding = gr.Slider(label='Only masked padding, pixels', minimum=0, maximum=256, step=4, value=32, elem_id="img2img_inpaint_full_res_padding")
 
-                            def select_img2img_tab(tab):
-                                return gr.update(visible=tab in [2, 3, 4]), gr.update(visible=tab == 3),
-
-                            for i, elem in enumerate(img2img_tabs):
-                                elem.select(
-                                    fn=lambda tab=i: select_img2img_tab(tab),
-                                    inputs=[],
-                                    outputs=[inpaint_controls, mask_alpha],
-                                )
-
                     if category not in {"accordions"}:
                         scripts.scripts_img2img.setup_ui_for_section(category)
 
-            img2img_gallery, generation_info, html_info, html_log = create_output_panel("img2img", opts.outdir_img2img_samples)
+            def select_img2img_tab(tab):
+                return gr.update(visible=tab in [2, 3, 4]), gr.update(visible=tab == 3),
+
+            for i, elem in enumerate(img2img_tabs):
+                elem.select(
+                    fn=lambda tab=i: select_img2img_tab(tab),
+                    inputs=[],
+                    outputs=[inpaint_controls, mask_alpha],
+                )
+
+            img2img_gallery, generation_info, html_info, html_log = create_output_panel("img2img", opts.outdir_img2img_samples, toprow)
 
             img2img_args = dict(
                 fn=wrap_gradio_gpu_call(modules.img2img.img2img, extra_outputs=[None, '', '']),
@@ -1296,7 +1229,7 @@ def create_ui():
 
             loadsave.setup_ui()
 
-        if os.path.exists(os.path.join(script_path, "notification.mp3")):
+        if os.path.exists(os.path.join(script_path, "notification.mp3")) and shared.opts.notification_audio:
             gr.Audio(interactive=False, value=os.path.join(script_path, "notification.mp3"), elem_id="audio_notification", visible=False)
 
         footer = shared.html("footer.html")

modules/ui_common.py

@@ -104,7 +104,7 @@ def save_files(js_data, images, do_make_zip, index):
     return gr.File.update(value=fullfns, visible=True), plaintext_to_html(f"Saved: {filenames[0]}")
 
 
-def create_output_panel(tabname, outdir):
+def create_output_panel(tabname, outdir, toprow=None):
 
     def open_folder(f):
         if not os.path.exists(f):
@@ -130,12 +130,15 @@ Requested path was: {f}
             else:
                 sp.Popen(["xdg-open", path])
 
-    with gr.Column(variant='panel', elem_id=f"{tabname}_results"):
-        with gr.Group(elem_id=f"{tabname}_gallery_container"):
-            result_gallery = gr.Gallery(label='Output', show_label=False, elem_id=f"{tabname}_gallery", columns=4, preview=True, height=shared.opts.gallery_height or None)
+    with gr.Column(elem_id=f"{tabname}_results"):
+        if toprow:
+            toprow.create_inline_toprow_image()
 
-        generation_info = None
-        with gr.Column():
+        with gr.Column(variant='panel', elem_id=f"{tabname}_results_panel"):
+            with gr.Group(elem_id=f"{tabname}_gallery_container"):
+                result_gallery = gr.Gallery(label='Output', show_label=False, elem_id=f"{tabname}_gallery", columns=4, preview=True, height=shared.opts.gallery_height or None)
+
+            generation_info = None
             with gr.Row(elem_id=f"image_buttons_{tabname}", elem_classes="image-buttons"):
                 open_folder_button = ToolButton(folder_symbol, elem_id=f'{tabname}_open_folder', visible=not shared.cmd_opts.hide_ui_dir_config, tooltip="Open images output directory.")
 

modules/ui_extra_networks.py

@@ -103,6 +103,7 @@ class ExtraNetworksPage:
         self.name = title.lower()
         self.id_page = self.name.replace(" ", "_")
         self.card_page = shared.html("extra-networks-card.html")
+        self.allow_prompt = True
         self.allow_negative_prompt = False
         self.metadata = {}
         self.items = {}
@@ -367,7 +368,7 @@ def create_ui(interface: gr.Blocks, unrelated_tabs, tabname):
     related_tabs = []
 
     for page in ui.stored_extra_pages:
-        with gr.Tab(page.title, id=page.id_page) as tab:
+        with gr.Tab(page.title, elem_id=f"{tabname}_{page.id_page}", elem_classes=["extra-page"]) as tab:
             elem_id = f"{tabname}_{page.id_page}_cards_html"
             page_elem = gr.HTML('Loading...', elem_id=elem_id)
             ui.pages.append(page_elem)
@@ -381,19 +382,28 @@ def create_ui(interface: gr.Blocks, unrelated_tabs, tabname):
             related_tabs.append(tab)
 
     edit_search = gr.Textbox('', show_label=False, elem_id=tabname+"_extra_search", elem_classes="search", placeholder="Search...", visible=False, interactive=True)
-    dropdown_sort = gr.Dropdown(choices=['Default Sort', 'Date Created', 'Date Modified', 'Name'], value='Default Sort', elem_id=tabname+"_extra_sort", elem_classes="sort", multiselect=False, visible=False, show_label=False, interactive=True, label=tabname+"_extra_sort_order")
-    button_sortorder = ToolButton(switch_values_symbol, elem_id=tabname+"_extra_sortorder", elem_classes="sortorder", visible=False, tooltip="Invert sort order")
+    dropdown_sort = gr.Dropdown(choices=['Name', 'Date Created', 'Date Modified', ], value=shared.opts.extra_networks_card_order_field, elem_id=tabname+"_extra_sort", elem_classes="sort", multiselect=False, visible=False, show_label=False, interactive=True, label=tabname+"_extra_sort_order")
+    button_sortorder = ToolButton(switch_values_symbol, elem_id=tabname+"_extra_sortorder", elem_classes=["sortorder"] + ([] if shared.opts.extra_networks_card_order == "Ascending" else ["sortReverse"]), visible=False, tooltip="Invert sort order")
     button_refresh = gr.Button('Refresh', elem_id=tabname+"_extra_refresh", visible=False)
     checkbox_show_dirs = gr.Checkbox(True, label='Show dirs', elem_id=tabname+"_extra_show_dirs", elem_classes="show-dirs", visible=False)
 
     ui.button_save_preview = gr.Button('Save preview', elem_id=tabname+"_save_preview", visible=False)
     ui.preview_target_filename = gr.Textbox('Preview save filename', elem_id=tabname+"_preview_filename", visible=False)
 
-    for tab in unrelated_tabs:
-        tab.select(fn=lambda: [gr.update(visible=False) for _ in range(5)], inputs=[], outputs=[edit_search, dropdown_sort, button_sortorder, button_refresh, checkbox_show_dirs], show_progress=False)
+    tab_controls = [edit_search, dropdown_sort, button_sortorder, button_refresh, checkbox_show_dirs]
 
-    for tab in related_tabs:
-        tab.select(fn=lambda: [gr.update(visible=True) for _ in range(5)], inputs=[], outputs=[edit_search, dropdown_sort, button_sortorder, button_refresh, checkbox_show_dirs], show_progress=False)
+    for tab in unrelated_tabs:
+        tab.select(fn=lambda: [gr.update(visible=False) for _ in tab_controls], _js='function(){ extraNetworksUrelatedTabSelected("' + tabname + '"); }', inputs=[], outputs=tab_controls, show_progress=False)
+
+    for page, tab in zip(ui.stored_extra_pages, related_tabs):
+        allow_prompt = "true" if page.allow_prompt else "false"
+        allow_negative_prompt = "true" if page.allow_negative_prompt else "false"
+
+        jscode = 'extraNetworksTabSelected("' + tabname + '", "' + f"{tabname}_{page.id_page}" + '", ' + allow_prompt + ', ' + allow_negative_prompt + ');'
+
+        tab.select(fn=lambda: [gr.update(visible=True) for _ in tab_controls],  _js='function(){ ' + jscode + ' }', inputs=[], outputs=tab_controls, show_progress=False)
+
+    dropdown_sort.change(fn=lambda: None, _js="function(){ applyExtraNetworkSort('" + tabname + "'); }")
 
     def pages_html():
         if not ui.pages_contents:

modules/ui_extra_networks_checkpoints.py

@@ -10,6 +10,8 @@ class ExtraNetworksPageCheckpoints(ui_extra_networks.ExtraNetworksPage):
     def __init__(self):
         super().__init__('Checkpoints')
 
+        self.allow_prompt = False
+
     def refresh(self):
         shared.refresh_checkpoints()
 

modules/ui_settings.py

@@ -1,6 +1,6 @@
 import gradio as gr
 
-from modules import ui_common, shared, script_callbacks, scripts, sd_models, sysinfo
+from modules import ui_common, shared, script_callbacks, scripts, sd_models, sysinfo, timer
 from modules.call_queue import wrap_gradio_call
 from modules.shared import opts
 from modules.ui_components import FormRow
@@ -177,8 +177,8 @@ class UiSettings:
                     download_localization = gr.Button(value='Download localization template', elem_id="download_localization")
                     reload_script_bodies = gr.Button(value='Reload custom script bodies (No ui updates, No restart)', variant='secondary', elem_id="settings_reload_script_bodies")
                     with gr.Row():
-                        unload_sd_model = gr.Button(value='Unload SD checkpoint to free VRAM', elem_id="sett_unload_sd_model")
-                        reload_sd_model = gr.Button(value='Reload the last SD checkpoint back into VRAM', elem_id="sett_reload_sd_model")
+                        unload_sd_model = gr.Button(value='Unload SD checkpoint to RAM', elem_id="sett_unload_sd_model")
+                        reload_sd_model = gr.Button(value='Load SD checkpoint to VRAM from RAM', elem_id="sett_reload_sd_model")
                     with gr.Row():
                         calculate_all_checkpoint_hash = gr.Button(value='Calculate hash for all checkpoint', elem_id="calculate_all_checkpoint_hash")
                         calculate_all_checkpoint_hash_threads = gr.Number(value=1, label="Number of parallel calculations", elem_id="calculate_all_checkpoint_hash_threads", precision=0, minimum=1)
@@ -194,16 +194,26 @@ class UiSettings:
 
                 self.text_settings = gr.Textbox(elem_id="settings_json", value=lambda: opts.dumpjson(), visible=False)
 
+            def call_func_and_return_text(func, text):
+                def handler():
+                    t = timer.Timer()
+                    func()
+                    t.record(text)
+
+                    return f'{text} in {t.total:.1f}s'
+
+                return handler
+
             unload_sd_model.click(
-                fn=sd_models.unload_model_weights,
+                fn=call_func_and_return_text(sd_models.unload_model_weights, 'Unloaded the checkpoint'),
                 inputs=[],
-                outputs=[]
+                outputs=[self.result]
             )
 
             reload_sd_model.click(
-                fn=sd_models.reload_model_weights,
+                fn=call_func_and_return_text(lambda: sd_models.send_model_to_device(shared.sd_model), 'Loaded the checkpoint'),
                 inputs=[],
-                outputs=[]
+                outputs=[self.result]
             )
 
             request_notifications.click(

modules/ui_toprow.py

@@ -0,0 +1,141 @@
+import gradio as gr
+
+from modules import shared, ui_prompt_styles
+import modules.images
+
+from modules.ui_components import ToolButton
+
+
+class Toprow:
+    """Creates a top row UI with prompts, generate button, styles, extra little buttons for things, and enables some functionality related to their operation"""
+
+    prompt = None
+    prompt_img = None
+    negative_prompt = None
+
+    button_interrogate = None
+    button_deepbooru = None
+
+    interrupt = None
+    skip = None
+    submit = None
+
+    paste = None
+    clear_prompt_button = None
+    apply_styles = None
+    restore_progress_button = None
+
+    token_counter = None
+    token_button = None
+    negative_token_counter = None
+    negative_token_button = None
+
+    ui_styles = None
+
+    submit_box = None
+
+    def __init__(self, is_img2img, is_compact=False):
+        id_part = "img2img" if is_img2img else "txt2img"
+        self.id_part = id_part
+        self.is_img2img = is_img2img
+        self.is_compact = is_compact
+
+        if not is_compact:
+            with gr.Row(elem_id=f"{id_part}_toprow", variant="compact"):
+                self.create_classic_toprow()
+        else:
+            self.create_submit_box()
+
+    def create_classic_toprow(self):
+        self.create_prompts()
+
+        with gr.Column(scale=1, elem_id=f"{self.id_part}_actions_column"):
+            self.create_submit_box()
+
+            self.create_tools_row()
+
+            self.create_styles_ui()
+
+    def create_inline_toprow_prompts(self):
+        if not self.is_compact:
+            return
+
+        self.create_prompts()
+
+        with gr.Row(elem_classes=["toprow-compact-stylerow"]):
+            with gr.Column(elem_classes=["toprow-compact-tools"]):
+                self.create_tools_row()
+            with gr.Column():
+                self.create_styles_ui()
+
+    def create_inline_toprow_image(self):
+        if not self.is_compact:
+            return
+
+        self.submit_box.render()
+
+    def create_prompts(self):
+        with gr.Column(elem_id=f"{self.id_part}_prompt_container", elem_classes=["prompt-container-compact"] if self.is_compact else [], scale=6):
+            with gr.Row(elem_id=f"{self.id_part}_prompt_row", elem_classes=["prompt-row"]):
+                self.prompt = gr.Textbox(label="Prompt", elem_id=f"{self.id_part}_prompt", show_label=False, lines=3, placeholder="Prompt (press Ctrl+Enter or Alt+Enter to generate)", elem_classes=["prompt"])
+                self.prompt_img = gr.File(label="", elem_id=f"{self.id_part}_prompt_image", file_count="single", type="binary", visible=False)
+
+            with gr.Row(elem_id=f"{self.id_part}_neg_prompt_row", elem_classes=["prompt-row"]):
+                self.negative_prompt = gr.Textbox(label="Negative prompt", elem_id=f"{self.id_part}_neg_prompt", show_label=False, lines=3, placeholder="Negative prompt (press Ctrl+Enter or Alt+Enter to generate)", elem_classes=["prompt"])
+
+        self.prompt_img.change(
+            fn=modules.images.image_data,
+            inputs=[self.prompt_img],
+            outputs=[self.prompt, self.prompt_img],
+            show_progress=False,
+        )
+
+    def create_submit_box(self):
+        with gr.Row(elem_id=f"{self.id_part}_generate_box", elem_classes=["generate-box"] + (["generate-box-compact"] if self.is_compact else []), render=not self.is_compact) as submit_box:
+            self.submit_box = submit_box
+
+            self.interrupt = gr.Button('Interrupt', elem_id=f"{self.id_part}_interrupt", elem_classes="generate-box-interrupt")
+            self.skip = gr.Button('Skip', elem_id=f"{self.id_part}_skip", elem_classes="generate-box-skip")
+            self.submit = gr.Button('Generate', elem_id=f"{self.id_part}_generate", variant='primary')
+
+            self.skip.click(
+                fn=lambda: shared.state.skip(),
+                inputs=[],
+                outputs=[],
+            )
+
+            self.interrupt.click(
+                fn=lambda: shared.state.interrupt(),
+                inputs=[],
+                outputs=[],
+            )
+
+    def create_tools_row(self):
+        with gr.Row(elem_id=f"{self.id_part}_tools"):
+            from modules.ui import paste_symbol, clear_prompt_symbol, restore_progress_symbol
+
+            self.paste = ToolButton(value=paste_symbol, elem_id="paste", tooltip="Read generation parameters from prompt or last generation if prompt is empty into user interface.")
+            self.clear_prompt_button = ToolButton(value=clear_prompt_symbol, elem_id=f"{self.id_part}_clear_prompt", tooltip="Clear prompt")
+            self.apply_styles = ToolButton(value=ui_prompt_styles.styles_materialize_symbol, elem_id=f"{self.id_part}_style_apply", tooltip="Apply all selected styles to prompts.")
+
+            if self.is_img2img:
+                self.button_interrogate = ToolButton('📎', tooltip='Interrogate CLIP - use CLIP neural network to create a text describing the image, and put it into the prompt field', elem_id="interrogate")
+                self.button_deepbooru = ToolButton('📦', tooltip='Interrogate DeepBooru - use DeepBooru neural network to create a text describing the image, and put it into the prompt field', elem_id="deepbooru")
+
+            self.restore_progress_button = ToolButton(value=restore_progress_symbol, elem_id=f"{self.id_part}_restore_progress", visible=False, tooltip="Restore progress")
+
+            self.token_counter = gr.HTML(value="<span>0/75</span>", elem_id=f"{self.id_part}_token_counter", elem_classes=["token-counter"])
+            self.token_button = gr.Button(visible=False, elem_id=f"{self.id_part}_token_button")
+            self.negative_token_counter = gr.HTML(value="<span>0/75</span>", elem_id=f"{self.id_part}_negative_token_counter", elem_classes=["token-counter"])
+            self.negative_token_button = gr.Button(visible=False, elem_id=f"{self.id_part}_negative_token_button")
+
+            self.clear_prompt_button.click(
+                fn=lambda *x: x,
+                _js="confirm_clear_prompt",
+                inputs=[self.prompt, self.negative_prompt],
+                outputs=[self.prompt, self.negative_prompt],
+            )
+
+    def create_styles_ui(self):
+        self.ui_styles = ui_prompt_styles.UiPromptStyles(self.id_part, self.prompt, self.negative_prompt)
+        self.ui_styles.setup_apply_button(self.apply_styles)

modules/xlmr_m18.py

@@ -0,0 +1,164 @@
+from transformers import BertPreTrainedModel,BertConfig
+import torch.nn as nn
+import torch
+from transformers.models.xlm_roberta.configuration_xlm_roberta import XLMRobertaConfig
+from transformers import XLMRobertaModel,XLMRobertaTokenizer
+from typing import Optional
+
+class BertSeriesConfig(BertConfig):
+    def __init__(self, vocab_size=30522, hidden_size=768, num_hidden_layers=12, num_attention_heads=12, intermediate_size=3072, hidden_act="gelu", hidden_dropout_prob=0.1, attention_probs_dropout_prob=0.1, max_position_embeddings=512, type_vocab_size=2, initializer_range=0.02, layer_norm_eps=1e-12, pad_token_id=0, position_embedding_type="absolute", use_cache=True, classifier_dropout=None,project_dim=512, pooler_fn="average",learn_encoder=False,model_type='bert',**kwargs):
+
+        super().__init__(vocab_size, hidden_size, num_hidden_layers, num_attention_heads, intermediate_size, hidden_act, hidden_dropout_prob, attention_probs_dropout_prob, max_position_embeddings, type_vocab_size, initializer_range, layer_norm_eps, pad_token_id, position_embedding_type, use_cache, classifier_dropout, **kwargs)
+        self.project_dim = project_dim
+        self.pooler_fn = pooler_fn
+        self.learn_encoder = learn_encoder
+
+class RobertaSeriesConfig(XLMRobertaConfig):
+    def __init__(self, pad_token_id=1, bos_token_id=0, eos_token_id=2,project_dim=512,pooler_fn='cls',learn_encoder=False, **kwargs):
+        super().__init__(pad_token_id=pad_token_id, bos_token_id=bos_token_id, eos_token_id=eos_token_id, **kwargs)
+        self.project_dim = project_dim
+        self.pooler_fn = pooler_fn
+        self.learn_encoder = learn_encoder
+
+
+class BertSeriesModelWithTransformation(BertPreTrainedModel):
+
+    _keys_to_ignore_on_load_unexpected = [r"pooler"]
+    _keys_to_ignore_on_load_missing = [r"position_ids", r"predictions.decoder.bias"]
+    config_class = BertSeriesConfig
+
+    def __init__(self, config=None, **kargs):
+        # modify initialization for autoloading
+        if config is None:
+            config = XLMRobertaConfig()
+            config.attention_probs_dropout_prob= 0.1
+            config.bos_token_id=0
+            config.eos_token_id=2
+            config.hidden_act='gelu'
+            config.hidden_dropout_prob=0.1
+            config.hidden_size=1024
+            config.initializer_range=0.02
+            config.intermediate_size=4096
+            config.layer_norm_eps=1e-05
+            config.max_position_embeddings=514
+
+            config.num_attention_heads=16
+            config.num_hidden_layers=24
+            config.output_past=True
+            config.pad_token_id=1
+            config.position_embedding_type= "absolute"
+
+            config.type_vocab_size= 1
+            config.use_cache=True
+            config.vocab_size= 250002
+            config.project_dim = 1024
+            config.learn_encoder = False
+        super().__init__(config)
+        self.roberta = XLMRobertaModel(config)
+        self.transformation = nn.Linear(config.hidden_size,config.project_dim)
+        # self.pre_LN=nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
+        self.tokenizer = XLMRobertaTokenizer.from_pretrained('xlm-roberta-large')
+        # self.pooler = lambda x: x[:,0]
+        # self.post_init()
+
+        self.has_pre_transformation = True
+        if self.has_pre_transformation:
+            self.transformation_pre = nn.Linear(config.hidden_size, config.project_dim)
+            self.pre_LN = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
+        self.post_init()
+
+    def encode(self,c):
+        device = next(self.parameters()).device
+        text = self.tokenizer(c,
+                        truncation=True,
+                        max_length=77,
+                        return_length=False,
+                        return_overflowing_tokens=False,
+                        padding="max_length",
+                        return_tensors="pt")
+        text["input_ids"] = torch.tensor(text["input_ids"]).to(device)
+        text["attention_mask"] = torch.tensor(
+            text['attention_mask']).to(device)
+        features = self(**text)
+        return features['projection_state']
+
+    def forward(
+        self,
+        input_ids: Optional[torch.Tensor] = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        token_type_ids: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.Tensor] = None,
+        head_mask: Optional[torch.Tensor] = None,
+        inputs_embeds: Optional[torch.Tensor] = None,
+        encoder_hidden_states: Optional[torch.Tensor] = None,
+        encoder_attention_mask: Optional[torch.Tensor] = None,
+        output_attentions: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+    ) :
+        r"""
+        """
+
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+
+        outputs = self.roberta(
+            input_ids=input_ids,
+            attention_mask=attention_mask,
+            token_type_ids=token_type_ids,
+            position_ids=position_ids,
+            head_mask=head_mask,
+            inputs_embeds=inputs_embeds,
+            encoder_hidden_states=encoder_hidden_states,
+            encoder_attention_mask=encoder_attention_mask,
+            output_attentions=output_attentions,
+            output_hidden_states=True,
+            return_dict=return_dict,
+        )
+
+        # # last module outputs
+        # sequence_output = outputs[0]
+
+
+        # # project every module
+        # sequence_output_ln = self.pre_LN(sequence_output)
+
+        # # pooler
+        # pooler_output = self.pooler(sequence_output_ln)
+        # pooler_output = self.transformation(pooler_output)
+        # projection_state = self.transformation(outputs.last_hidden_state)
+
+        if self.has_pre_transformation:
+            sequence_output2 = outputs["hidden_states"][-2]
+            sequence_output2 = self.pre_LN(sequence_output2)
+            projection_state2 = self.transformation_pre(sequence_output2)
+
+            return {
+                "projection_state": projection_state2,
+                "last_hidden_state": outputs.last_hidden_state,
+                "hidden_states": outputs.hidden_states,
+                "attentions": outputs.attentions,
+            }
+        else:
+            projection_state = self.transformation(outputs.last_hidden_state)
+            return {
+                "projection_state": projection_state,
+                "last_hidden_state": outputs.last_hidden_state,
+                "hidden_states": outputs.hidden_states,
+                "attentions": outputs.attentions,
+            }
+
+
+        # return {
+        #     'pooler_output':pooler_output,
+        #     'last_hidden_state':outputs.last_hidden_state,
+        #     'hidden_states':outputs.hidden_states,
+        #     'attentions':outputs.attentions,
+        #     'projection_state':projection_state,
+        #     'sequence_out': sequence_output
+        # }
+
+
+class RobertaSeriesModelWithTransformation(BertSeriesModelWithTransformation):
+    base_model_prefix = 'roberta'
+    config_class= RobertaSeriesConfig

requirements.txt

@@ -32,3 +32,5 @@ torch
 torchdiffeq
 torchsde
 transformers==4.30.2
+
+git+https://github.com/openai/consistencydecoder.git

requirements_versions.txt

@@ -29,3 +29,5 @@ torch
 torchdiffeq==0.2.3
 torchsde==0.2.6
 transformers==4.30.2
+httpx==0.24.1
+git+https://github.com/openai/consistencydecoder.git

script.js

@@ -124,16 +124,20 @@ document.addEventListener("DOMContentLoaded", function() {
  * Add a ctrl+enter as a shortcut to start a generation
  */
 document.addEventListener('keydown', function(e) {
-    var handled = false;
-    if (e.key !== undefined) {
-        if ((e.key == "Enter" && (e.metaKey || e.ctrlKey || e.altKey))) handled = true;
-    } else if (e.keyCode !== undefined) {
-        if ((e.keyCode == 13 && (e.metaKey || e.ctrlKey || e.altKey))) handled = true;
-    }
-    if (handled) {
-        var button = get_uiCurrentTabContent().querySelector('button[id$=_generate]');
-        if (button) {
-            button.click();
+    const isEnter = e.key === 'Enter' || e.keyCode === 13;
+    const isModifierKey = e.metaKey || e.ctrlKey || e.altKey;
+
+    const interruptButton = get_uiCurrentTabContent().querySelector('button[id$=_interrupt]');
+    const generateButton = get_uiCurrentTabContent().querySelector('button[id$=_generate]');
+
+    if (isEnter && isModifierKey) {
+        if (interruptButton.style.display === 'block') {
+            interruptButton.click();
+            setTimeout(function() {
+                generateButton.click();
+            }, 500);
+        } else {
+            generateButton.click();
         }
         e.preventDefault();
     }

scripts/prompts_from_file.py

@@ -114,6 +114,7 @@ class Script(scripts.Script):
     def ui(self, is_img2img):
         checkbox_iterate = gr.Checkbox(label="Iterate seed every line", value=False, elem_id=self.elem_id("checkbox_iterate"))
         checkbox_iterate_batch = gr.Checkbox(label="Use same random seed for all lines", value=False, elem_id=self.elem_id("checkbox_iterate_batch"))
+        prompt_position = gr.Radio(["start", "end"], label="Insert prompts at the", elem_id=self.elem_id("prompt_position"), value="start")
 
         prompt_txt = gr.Textbox(label="List of prompt inputs", lines=1, elem_id=self.elem_id("prompt_txt"))
         file = gr.File(label="Upload prompt inputs", type='binary', elem_id=self.elem_id("file"))
@@ -124,9 +125,9 @@ class Script(scripts.Script):
         # We don't shrink back to 1, because that causes the control to ignore [enter], and it may
         # be unclear to the user that shift-enter is needed.
         prompt_txt.change(lambda tb: gr.update(lines=7) if ("\n" in tb) else gr.update(lines=2), inputs=[prompt_txt], outputs=[prompt_txt], show_progress=False)
-        return [checkbox_iterate, checkbox_iterate_batch, prompt_txt]
+        return [checkbox_iterate, checkbox_iterate_batch, prompt_position, prompt_txt]
 
-    def run(self, p, checkbox_iterate, checkbox_iterate_batch, prompt_txt: str):
+    def run(self, p, checkbox_iterate, checkbox_iterate_batch, prompt_position, prompt_txt: str):
         lines = [x for x in (x.strip() for x in prompt_txt.splitlines()) if x]
 
         p.do_not_save_grid = True
@@ -167,6 +168,18 @@ class Script(scripts.Script):
                 else:
                     setattr(copy_p, k, v)
 
+            if args.get("prompt") and p.prompt:
+                if prompt_position == "start":
+                    copy_p.prompt = args.get("prompt") + " " + p.prompt
+                else:
+                    copy_p.prompt = p.prompt + " " + args.get("prompt")
+
+            if args.get("negative_prompt") and p.negative_prompt:
+                if prompt_position == "start":
+                    copy_p.negative_prompt = args.get("negative_prompt") + " " + p.negative_prompt
+                else:
+                    copy_p.negative_prompt = p.negative_prompt + " " + args.get("negative_prompt")
+
             proc = process_images(copy_p)
             images += proc.images
 

style.css

@@ -204,6 +204,11 @@ div.block.gradio-accordion {
     padding: 8px 8px;
 }
 
+input[type="checkbox"].input-accordion-checkbox{
+    vertical-align: sub;
+    margin-right: 0.5em;
+}
+
 
 /* txt2img/img2img specific */
 
@@ -291,6 +296,13 @@ div.block.gradio-accordion {
     min-height: 4.5em;
 }
 
+#txt2img_generate, #img2img_generate {
+    min-height: 4.5em;
+}
+.generate-box-compact #txt2img_generate, .generate-box-compact #img2img_generate {
+    min-height: 3em;
+}
+
 @media screen and (min-width: 2500px) {
     #txt2img_gallery, #img2img_gallery {
         min-height: 768px;
@@ -398,6 +410,15 @@ div#extras_scale_to_tab div.form{
     min-width: 0.5em;
 }
 
+div.toprow-compact-stylerow{
+    margin: 0.5em 0;
+}
+
+div.toprow-compact-tools{
+    min-width: fit-content !important;
+    max-width: fit-content;
+}
+
 /* settings */
 #quicksettings {
     align-items: end;
@@ -520,7 +541,8 @@ table.popup-table .link{
     height: 20px;
     background: #b4c0cc;
     border-radius: 3px !important;
-    top: -20px;
+    top: -14px;
+    left: 0px;
     width: 100%;
 }
 
@@ -818,6 +840,10 @@ footer {
 
 /* extra networks UI */
 
+.extra-page .prompt{
+    margin: 0 0 0.5em 0;
+}
+
 .extra-network-cards{
     height: calc(100vh - 24rem);
     overflow: clip scroll;

webui.bat

@@ -1,5 +1,9 @@
 @echo off
 
+if exist webui.settings.bat (
+    call webui.settings.bat
+)
+
 if not defined PYTHON (set PYTHON=python)
 if defined GIT (set "GIT_PYTHON_GIT_EXECUTABLE=%GIT%")
 if not defined VENV_DIR (set "VENV_DIR=%~dp0%venv")