Fix - reduce the amount of VRAM occupied when the program starts up, this caused a regression and failures on GPUs with 4 gb or less of VRAM

Merge pull request #485 from cmdr2/beta
UI setting for preventing browser autostart
2022-11-16 19:29:04 +05:30 · 2022-11-16 12:45:51 +05:30 · 2022-11-16 12:43:46 +05:30 · 2022-11-16 12:29:09 +05:30 · 2022-11-16 12:20:25 +05:30 · 2022-11-16 12:16:46 +05:30
19 changed files with 2703 additions and 919 deletions
--- a/scripts/on_env_start.bat
+++ b/scripts/on_env_start.bat
@ -50,7 +50,7 @@ if "%update_branch%"=="" (
    )
 )

-@xcopy sd-ui-files\ui ui /s /i /Y
+@xcopy sd-ui-files\ui ui /s /i /Y /q
@copy sd-ui-files\scripts\on_sd_start.bat scripts\ /Y
@copy sd-ui-files\scripts\bootstrap.bat scripts\ /Y
@copy "sd-ui-files\scripts\Start Stable Diffusion UI.cmd" . /Y
--- a/scripts/on_sd_start.bat
+++ b/scripts/on_sd_start.bat
@ -199,7 +199,9 @@ call WHERE uvicorn > .tmp


 if not exist "..\models\stable-diffusion" mkdir "..\models\stable-diffusion"
+if not exist "..\models\vae" mkdir "..\models\vae"
 echo. > "..\models\stable-diffusion\Put your custom ckpt files here.txt"
+echo. > "..\models\vae\Put your VAE files here.txt"

@if exist "sd-v1-4.ckpt" (
    for %%I in ("sd-v1-4.ckpt") do if "%%~zI" EQU "4265380512" (
@ -329,6 +331,36 @@ echo. > "..\models\stable-diffusion\Put your custom ckpt files here.txt"



+@if exist "..\models\vae\vae-ft-mse-840000-ema-pruned.ckpt" (
+    for %%I in ("..\models\vae\vae-ft-mse-840000-ema-pruned.ckpt") do if "%%~zI" EQU "334695179" (
+        echo "Data files (weights) necessary for the default VAE (sd-vae-ft-mse-original) were already downloaded"
+    ) else (
+        echo. & echo "The default VAE (sd-vae-ft-mse-original) file present at models\vae\vae-ft-mse-840000-ema-pruned.ckpt is invalid. It is only %%~zI bytes in size. Re-downloading.." & echo.
+        del "..\models\vae\vae-ft-mse-840000-ema-pruned.ckpt"
+    )
+)
+
+@if not exist "..\models\vae\vae-ft-mse-840000-ema-pruned.ckpt" (
+    @echo. & echo "Downloading data files (weights) for the default VAE (sd-vae-ft-mse-original).." & echo.
+
+    @call curl -L -k https://huggingface.co/stabilityai/sd-vae-ft-mse-original/resolve/main/vae-ft-mse-840000-ema-pruned.ckpt > ..\models\vae\vae-ft-mse-840000-ema-pruned.ckpt
+
+    @if exist "..\models\vae\vae-ft-mse-840000-ema-pruned.ckpt" (
+        for %%I in ("..\models\vae\vae-ft-mse-840000-ema-pruned.ckpt") do if "%%~zI" NEQ "334695179" (
+            echo. & echo "Error: The downloaded default VAE (sd-vae-ft-mse-original) file was invalid! Bytes downloaded: %%~zI" & echo.
+            echo. & echo "Error downloading the data files (weights) for the default VAE (sd-vae-ft-mse-original). Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!" & echo.
+            pause
+            exit /b
+        )
+    ) else (
+        @echo. & echo "Error downloading the data files (weights) for the default VAE (sd-vae-ft-mse-original). Sorry about that, please try to:" & echo "  1. Run this installer again." & echo "  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" & echo "  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB" & echo "  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues" & echo "Thanks!" & echo.
+        pause
+        exit /b
+    )
+)
+
+
+
@>nul findstr /m "sd_install_complete" ..\scripts\install_status.txt
@if "%ERRORLEVEL%" NEQ "0" (
    @echo sd_weights_downloaded >> ..\scripts\install_status.txt
--- a/scripts/on_sd_start.sh
+++ b/scripts/on_sd_start.sh
@ -159,7 +159,9 @@ fi


 mkdir -p "../models/stable-diffusion"
+mkdir -p "../models/vae"
 echo "" > "../models/stable-diffusion/Put your custom ckpt files here.txt"
+echo "" > "../models/vae/Put your VAE files here.txt"

 if [ -f "sd-v1-4.ckpt" ]; then
    model_size=`find "sd-v1-4.ckpt" -printf "%s"`
@ -269,6 +271,38 @@ if [ ! -f "RealESRGAN_x4plus_anime_6B.pth" ]; then
 fi


+if [ -f "../models/vae/vae-ft-mse-840000-ema-pruned.ckpt" ]; then
+    model_size=`find ../models/vae/vae-ft-mse-840000-ema-pruned.ckpt -printf "%s"`
+
+    if [ "$model_size" -eq "334695179" ]; then
+        echo "Data files (weights) necessary for the default VAE (sd-vae-ft-mse-original) were already downloaded"
+    else
+        printf "\n\nThe model file present at models/vae/vae-ft-mse-840000-ema-pruned.ckpt is invalid. It is only $model_size bytes in size. Re-downloading.."
+        rm ../models/vae/vae-ft-mse-840000-ema-pruned.ckpt
+    fi
+fi
+
+if [ ! -f "../models/vae/vae-ft-mse-840000-ema-pruned.ckpt" ]; then
+    echo "Downloading data files (weights) for the default VAE (sd-vae-ft-mse-original).."
+
+    curl -L -k https://huggingface.co/stabilityai/sd-vae-ft-mse-original/resolve/main/vae-ft-mse-840000-ema-pruned.ckpt > ../models/vae/vae-ft-mse-840000-ema-pruned.ckpt
+
+    if [ -f "../models/vae/vae-ft-mse-840000-ema-pruned.ckpt" ]; then
+        model_size=`find ../models/vae/vae-ft-mse-840000-ema-pruned.ckpt -printf "%s"`
+        if [ ! "$model_size" -eq "334695179" ]; then
+            printf "\n\nError: The downloaded default VAE (sd-vae-ft-mse-original) file was invalid! Bytes downloaded: $model_size\n\n"
+            printf "\n\nError downloading the data files (weights) for the default VAE (sd-vae-ft-mse-original). Sorry about that, please try to:\n  1. Run this installer again.\n  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting\n  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB\n  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues\nThanks!\n\n"
+            read -p "Press any key to continue"
+            exit
+        fi
+    else
+        printf "\n\nError downloading the data files (weights) for the default VAE (sd-vae-ft-mse-original). Sorry about that, please try to:\n  1. Run this installer again.\n  2. If that doesn't fix it, please try the common troubleshooting steps at https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting\n  3. If those steps don't help, please copy *all* the error messages in this window, and ask the community at https://discord.com/invite/u9yhsFmEkB\n  4. If that doesn't solve the problem, please file an issue at https://github.com/cmdr2/stable-diffusion-ui/issues\nThanks!\n\n"
+        read -p "Press any key to continue"
+        exit
+    fi
+fi
+
+
 if [ `grep -c sd_install_complete ../scripts/install_status.txt` -gt "0" ]; then
    echo sd_weights_downloaded >> ../scripts/install_status.txt
    echo sd_install_complete >> ../scripts/install_status.txt
--- a/ui/index.html
+++ b/ui/index.html
@ -1,14 +1,15 @@
 <!DOCTYPE html>
 <html>
 <head>
+    <title>Stable Diffusion UI</title>
    <meta name="viewport" content="width=device-width, initial-scale=1.0">
    <link rel="icon" type="image/png" href="/media/images/favicon-16x16.png" sizes="16x16">
    <link rel="icon" type="image/png" href="/media/images/favicon-32x32.png" sizes="32x32">
    <link rel="stylesheet" href="/media/css/fonts.css?v=1">
-    <link rel="stylesheet" href="/media/css/themes.css?v=1">
-    <link rel="stylesheet" href="/media/css/main.css?v=3">
-    <link rel="stylesheet" href="/media/css/auto-save.css?v=2">
-    <link rel="stylesheet" href="/media/css/modifier-thumbnails.css?v=2">
+    <link rel="stylesheet" href="/media/css/themes.css?v=3">
+    <link rel="stylesheet" href="/media/css/main.css?v=17">
+    <link rel="stylesheet" href="/media/css/auto-save.css?v=5">
+    <link rel="stylesheet" href="/media/css/modifier-thumbnails.css?v=4">
    <link rel="stylesheet" href="/media/css/fontawesome-all.min.css?v=1">
    <link rel="stylesheet" href="/media/css/drawingboard.min.css">
    <script src="/media/js/jquery-3.6.1.min.js"></script>
@ -18,59 +19,39 @@
 <div id="container">
    <div id="top-nav">
        <div id="logo">
-            <h1>Stable Diffusion UI <small>v2.3.5 <span id="updateBranchLabel"></span></small></h1>
+            <h1>Stable Diffusion UI <small>v2.4.6 <span id="updateBranchLabel"></span></small></h1>
        </div>
-        <ul id="top-nav-items">
-            <li class="dropdown">
-                <span><i class="fa fa-comments icon"></i> Help & Community</span>
-                <ul id="community-links" class="dropdown-content">
-                    <li><a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/" target="_blank"><i class="fa-solid fa-book fa-fw"></i> User guide</a></li>
-                    <li><a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" target="_blank"><i class="fa-solid fa-circle-question fa-fw"></i> Usual problems and solutions</a></li>
-                    <li><a href="https://discord.com/invite/u9yhsFmEkB" target="_blank"><i class="fa-brands fa-discord fa-fw"></i> Discord user community</a></li>
-                    <li><a href="https://www.reddit.com/r/StableDiffusionUI/" target="_blank"><i class="fa-brands fa-reddit fa-fw"></i> Reddit community</a></li>
-                    <li><a href="https://github.com/cmdr2/stable-diffusion-ui" target="_blank"><i class="fa-brands fa-github fa-fw"></i> Source code on GitHub</a></li>
-                </ul>
-            </li>
-            <li class="dropdown">
+        <div id="server-status">
+            <div id="server-status-color">●</div>
+            <span id="server-status-msg">Stable Diffusion is starting..</span>
+        </div>
+        <div id="tab-container">
+            <span id="tab-main" class="tab active">
+                <span><i class="fa fa-image icon"></i> Generate</span>
+            </span>
+            <span id="tab-settings" class="tab">
                <span><i class="fa fa-gear icon"></i> Settings</span>
-                <div id="system-settings" class="panel-box settings-box dropdown-content">
-                    <ul id="system-settings-entries">
-                        <li><b class="settings-subheader">System Settings</b></li>
-                        <br/>
-                        <li><label for="theme">Theme: </label><select id="theme" name="theme"><option value="theme-default">Default</option></select></li>
-                        <li><input id="save_to_disk" name="save_to_disk" type="checkbox"> <label for="save_to_disk">Automatically save to <input id="diskPath" name="diskPath" size="40" disabled></label></li>
-                        <li><input id="sound_toggle" name="sound_toggle" type="checkbox" checked> <label for="sound_toggle">Play sound on task completion</label></li>
-                        <li><input id="turbo" name="turbo" type="checkbox" checked> <label for="turbo">Turbo mode <small>(generates images faster, but uses an additional 1 GB of GPU memory)</small></label></li>
-                        <li><input id="use_cpu" name="use_cpu" type="checkbox"> <label for="use_cpu">Use CPU instead of GPU <small>(warning: this will be *very* slow)</small></label></li>
-                        <li><input id="use_full_precision" name="use_full_precision" type="checkbox"> <label for="use_full_precision">Use full precision <small>(for GPU-only. warning: this will consume more VRAM)</small></label></li>
-                        <li>
-                            <input id="auto_save_settings" name="auto_save_settings" checked type="checkbox">
-                            <label for="auto_save_settings">Automatically save settings <small>(settings restored on browser load)</small></label>
-                            <br/>
-                            <button id="configureSettingsSaveBtn">Configure</button>
-                        </li>
-                        <!-- <li><input id="allow_nsfw" name="allow_nsfw" type="checkbox"> <label for="allow_nsfw">Allow NSFW Content (You confirm you are above 18 years of age)</label></li> -->
-                        <br/>
-                        <li><input id="use_beta_channel" name="use_beta_channel" type="checkbox"> <label for="use_beta_channel">🔥Beta channel. Get the latest features immediately (but could be less stable). Please restart the program after changing this.</label></li>
-                    </ul>
-                </div>
-            </li>
-        </ul>
+            </span>
+            <span id="tab-about" class="tab">
+                <span><i class="fa fa-comments icon"></i> Help & Community</span>
+            </span>
+        </div>
    </div>

-    <div class="flex-container">
-        <div id="editor" class="col-fixed-10">
-            <div id="server-status">
-                <div id="server-status-color">●</div>
-                <span id="server-status-msg">Stable Diffusion is starting..</span>
-            </div>
+    <div id="tab-content-wrapper">
+        <div id="tab-content-main" class="tab-content active flex-container">
+        <div id="editor">
            <div id="editor-inputs">
                <div id="editor-inputs-prompt" class="row">
                    <label for="prompt"><b>Enter Prompt</b></label> <small>or</small> <button id="promptsFromFileBtn">Load from a file</button>
                    <textarea id="prompt" class="col-free">a photograph of an astronaut riding a horse</textarea>
                    <input id="prompt_from_file" name="prompt_from_file" type="file" /> <!-- hidden -->

-                    <label for="negative_prompt" class="collapsible" id="negative_prompt_handle">Negative Prompt <small>(optional)</small></label>
+                    <label for="negative_prompt" class="collapsible" id="negative_prompt_handle">
+                        Negative Prompt
+                        <a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/Writing-prompts#negative-prompts" target="_blank"><i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip right">Click to learn more about Negative Prompts</span></i></a>
+                        <small>(optional)</small>
+                    </label>
                    <div class="collapsible-content">
                        <input id="negative_prompt" name="negative_prompt" placeholder="list the things to remove from the image (e.g. fog, green)">
                    </div>
@ -87,7 +68,12 @@
                        </div>

                        <br/>
-                        <input id="enable_mask" name="enable_mask" type="checkbox"> <label for="enable_mask">In-Painting (beta) <small>(select the area which the AI will paint into)</small></label>
+                        <input id="enable_mask" name="enable_mask" type="checkbox">
+                        <label for="enable_mask">
+                            In-Painting (beta)
+                            <a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/Inpainting" target="_blank"><i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip right">Click to learn more about InPainting</span></i></a>
+                            <small>(select the area which the AI will paint into)</small>
+                        </label>
                        <div id="inpaintingEditor"></div>
                    </div>
                </div>
@ -101,19 +87,19 @@
                <button id="stopImage" class="secondaryButton">Stop All</button>
            </div>

-            <div class="line-separator">&nbsp;</div>
+            <span class="line-separator"></span>

-            <div id="editor-settings" class="panel-box settings-box">
+            <div id="editor-settings" class="settings-box panel-box">
                <h4 class="collapsible">
                    Image Settings
-                    <i id="reset-image-settings" class="fa-solid fa-arrow-rotate-left">
-                        <span class="simple-tooltip right">
+                    <i id="reset-image-settings" class="fa-solid fa-arrow-rotate-left section-button">
+                        <span class="simple-tooltip left">
                            Reset Image Settings
                        </span>
                    </i>
                </h4>
-                <ul id="editor-settings-entries" class="collapsible-content">
-                    <li><table>
+                <div id="editor-settings-entries" class="collapsible-content">
+                    <div><table>
                    <tr><b class="settings-subheader">Image Settings</b></tr>
                    <tr class="pl-5"><td><label for="seed">Seed:</label></td><td><input id="seed" name="seed" size="10" value="30000" onkeypress="preventNonNumericalInput(event)"> <input id="random_seed" name="random_seed" type="checkbox" checked><label for="random_seed">Random</label></td></tr>
                    <tr class="pl-5"><td><label for="num_outputs_total">Number of Images:</label></td><td><input id="num_outputs_total" name="num_outputs_total" value="1" size="1" onkeypress="preventNonNumericalInput(event)"> <label><small>(total)</small></label> <input id="num_outputs_parallel" name="num_outputs_parallel" value="1" size="1" onkeypress="preventNonNumericalInput(event)"> <label for="num_outputs_parallel"><small>(in parallel)</small></label></td></tr>
@ -121,6 +107,13 @@
                        <select id="stable_diffusion_model" name="stable_diffusion_model">
                            <!-- <option value="sd-v1-4" selected>sd-v1-4</option> -->
                        </select>
+                        <a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/Custom-Models" target="_blank"><i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip right">Click to learn more about custom models</span></i></a>
+                    </td></tr>
+                    <tr class="pl-5"><td><label for="vae_model">Custom VAE:</i></label></td><td>
+                        <select id="vae_model" name="vae_model">
+                            <!-- <option value="" selected>None</option> -->
+                        </select>
+                        <a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/VAE-Variational-Auto-Encoder" target="_blank"><i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip right">Click to learn more about VAEs</span></i></a>
                    </td></tr>
                    <tr id="samplerSelection" class="pl-5"><td><label for="sampler">Sampler:</label></td><td>
                        <select id="sampler" name="sampler">
@ -133,6 +126,7 @@
                            <option value="dpm2_a">dpm2_a</option>
                            <option value="lms">lms</option>
                        </select>
+                        <a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/How-to-Use#samplers" target="_blank"><i class="fa-solid fa-circle-question help-btn"><span class="simple-tooltip right">Click to learn more about samplers</span></i></a>
                    </td></tr>
                    <tr class="pl-5"><td><label>Image Size: </label></td><td>
                        <select id="width" name="width" value="512">
@ -189,12 +183,11 @@
                            <option value="png">png</option>
                        </select>
                    </td></tr>
-                    </li></table>
-
-                    <br/>
+                    </table></div>

+                    <div><ul>
                    <li><b class="settings-subheader">Render Settings</b></li>
-                    <li class="pl-5"><input id="stream_image_progress" name="stream_image_progress" type="checkbox"> <label for="stream_image_progress">Show a live preview <small>(uses more VRAM, slightly slower image creation)</small></label></li>
+                    <li class="pl-5"><input id="stream_image_progress" name="stream_image_progress" type="checkbox"> <label for="stream_image_progress">Show a live preview <small>(uses more VRAM, and slower image creation)</small></label></li>
                    <li class="pl-5"><input id="use_face_correction" name="use_face_correction" type="checkbox"> <label for="use_face_correction">Fix incorrect faces and eyes <small>(uses GFPGAN)</small></label></li>
                    <li class="pl-5">
                        <input id="use_upscale" name="use_upscale" type="checkbox"> <label for="use_upscale">Upscale image by 4x with </label>
@ -204,12 +197,19 @@
                        </select>
                    </li>
                    <li class="pl-5"><input id="show_only_filtered_image" name="show_only_filtered_image" type="checkbox" checked> <label for="show_only_filtered_image">Show only the corrected/upscaled image</label></li>
-                </ul>
+                    </ul></div>
+                </div>
            </div>

            <div id="editor-modifiers" class="panel-box">
-                <button id="modifier-settings-btn" title="Add custom modifiers"><i class="fa fa-gear"></i></button>
-                <h4 class="collapsible">Image Modifiers (art styles, tags etc)</h4>
+                <h4 class="collapsible">
+                    Image Modifiers (art styles, tags etc)
+                    <i id="modifier-settings-btn" class="fa-solid fa-gear section-button">
+                        <span class="simple-tooltip left">
+                            Add Custom Modifiers
+                        </span>
+                    </i>
+                </h4>
                <div id="editor-modifiers-entries" class="collapsible-content">
                    <div id="editor-modifiers-entries-toolbar">
                        <label for="preview-image">Image Style:</label>
@ -236,21 +236,76 @@
                <button id="clear-all-previews" class="secondaryButton"><i class="fa-solid fa-trash-can"></i> Clear All</button>
            </div>
        </div>
-    </div>
+        </div>

-    <div id="save-settings-config" style="display:none">
+        <div id="tab-content-settings" class="tab-content">
+            <div id="system-settings" class="tab-content-inner">
+                <h1>System Settings</h1>
+                <table class="form-table"></table>
+                <br/>
+                <button id="save-system-settings-btn" class="primaryButton">Save</button>
+                <br/><br/>
+                <div>
+                    <h3><i class="fa fa-microchip icon"></i> System Info</h3>
+                    <div id="system-info"></div>
+                </div>
+            </div>
+        </div>
+        <div id="tab-content-about" class="tab-content">
+            <div class="tab-content-inner">
+                <div class="float-container">
+                    <div class="float-child">
+                        <h1>Help</h1>
+                        <ul id="help-links">
+                            <li><span class="help-section">Using the software</span>
+                            <ul>
+                                <li> <a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/How-To-Use" target="_blank"><i class="fa-solid fa-book fa-fw"></i> How to use</a>
+                                <li> <a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/UI-Overview" target="_blank"><i class="fa-solid fa-list fa-fw"></i> UI Overview</a>
+                                <li> <a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/Writing-Prompts" target="_blank"><i class="fa-solid fa-pen-to-square fa-fw"></i> Writing prompts</a>
+                                <li> <a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/Inpainting" target="_blank"><i class="fa-solid fa-paintbrush fa-fw"></i> Inpainting</a>
+                            </ul>
+  
+                            <li><span class="help-section">Installation</span>
+                            <ul>
+                                <li> <a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/Troubleshooting" target="_blank"><i class="fa-solid fa-circle-question fa-fw"></i> Troubleshooting</a>
+                            </ul>
+  
+                            <li><span class="help-section">Downloadable Content</span>
+                            <ul>
+                                <li> <a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/Custom-Models" target="_blank"><i class="fa-solid fa-images fa-fw"></i> Custom Models</a>
+                                <li> <a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/UI-Plugins" target="_blank"><i class="fa-solid fa-puzzle-piece fa-fw"></i> UI Plugins</a>
+                                <li> <a href="https://github.com/cmdr2/stable-diffusion-ui/wiki/VAE-Variational-Auto-Encoder" target="_blank"><i class="fa-solid fa-hand-sparkles fa-fw"></i> VAE Variational Auto Encoder</a>
+                            </ul>
+                        </ul>
+                    </div>
+  
+                    <div class="float-child">
+                        <h1>Community</h1>
+                        <ul id="community-links">
+                            <li><a href="https://discord.com/invite/u9yhsFmEkB" target="_blank"><i class="fa-brands fa-discord fa-fw"></i> Discord user community</a></li>
+                            <li><a href="https://www.reddit.com/r/StableDiffusionUI/" target="_blank"><i class="fa-brands fa-reddit fa-fw"></i> Reddit community</a></li>
+                            <li><a href="https://github.com/cmdr2/stable-diffusion-ui" target="_blank"><i class="fa-brands fa-github fa-fw"></i> Source code on GitHub</a></li>
+                        </ul>
+                    </div>
+                </div>
+            </div>
+        </div>
+    </div>
+    
+
+    <div id="save-settings-config" class="popup">
        <div>
-            <span id="save-settings-config-close-btn">X</span>
+            <i class="close-button fa-solid fa-xmark"></i>
            <h1>Save Settings Configuration</h1>
            <p>Select which settings should be remembered when restarting the browser</p>
-            <table id="save-settings-config-table">
+            <table id="save-settings-config-table" class="form-table">
            </table>
        </div>
    </div>

-    <div id="modifier-settings-config" style="display:none">
+    <div id="modifier-settings-config" class="popup">
        <div>
-            <span id="modifier-settings-config-close-btn">X</span>
+            <i class="close-button fa-solid fa-xmark"></i>
            <h1>Modifier Settings</h1>
            <p>Set your custom modifiers (one per line)</p>
            <textarea id="custom-modifiers-input" placeholder="Enter your custom modifiers, one-per-line"></textarea>
@ -258,9 +313,9 @@
        </div>
    </div>

-    <div class="line-separator">&nbsp;</div>
-
-    <div id="footer" class="panel-box">
+    <div id="footer-spacer"></div>
+    <div id="footer">
+        <div class="line-separator">&nbsp;</div>
        <p>If you found this project useful and want to help keep it alive, please <a href="https://ko-fi.com/cmdr2_stablediffusion_ui" target="_blank"><img src="/media/images/kofi.png" id="coffeeButton"></a> to help cover the cost of development and maintenance! Thank you for your support!</p>
        <p>Please feel free to join the <a href="https://discord.com/invite/u9yhsFmEkB" target="_blank">discord community</a> or <a href="https://github.com/cmdr2/stable-diffusion-ui/issues" target="_blank">file an issue</a> if you have any problems or suggestions in using this interface.</p>
        <div id="footer-legal">
@ -272,13 +327,15 @@
 </div>
 </body>

+<script src="media/js/parameters.js?v=9"></script>
 <script src="media/js/plugins.js?v=1"></script>
-<script src="media/js/utils.js?v=5"></script>
+<script src="media/js/utils.js?v=6"></script>
 <script src="media/js/inpainting-editor.js?v=1"></script>
-<script src="media/js/image-modifiers.js?v=3"></script>
-<script src="media/js/auto-save.js?v=2.3"></script>
-<script src="media/js/main.js?v=5"></script>
-<script src="media/js/themes.js?v=2"></script>
+<script src="media/js/image-modifiers.js?v=6"></script>
+<script src="media/js/auto-save.js?v=8"></script>
+<script src="media/js/main.js?v=22.1"></script>
+<script src="media/js/themes.js?v=4"></script>
+<script src="media/js/dnd.js?v=9"></script>
 <script>
 async function init() {
    await initSettings()
@ -287,6 +344,7 @@ async function init() {
    await getAppConfig()
    await loadModifiers()
    await loadUIPlugins()
+    await getDevices()

    setInterval(healthCheck, HEALTH_PING_INTERVAL * 1000)
    healthCheck()
--- a/ui/media/css/auto-save.css
+++ b/ui/media/css/auto-save.css
@ -6,69 +6,43 @@
    display: none;
 }

-#save-settings-config {
-    position: absolute;
-    background: rgba(32, 33, 36, 50%);
-    top: 0px;
-    left: 0px;
-    right: 0px;
-    bottom: 0px;
-    z-index: 1000;
-}
-
-#save-settings-config > div {
-    background: var(--background-color3);
-    max-width: 600px;
-    margin: auto;
-    margin-top: 50px;
-    border-radius: 6px;
-    padding: 30px;
-    text-align: center;
-}
-
-#save-settings-config-table {
+.form-table {
    margin: auto;
 }

-#save-settings-config-table th {
+.form-table th {
    padding-top: 15px;
    padding-bottom: 5px;
 }

-#save-settings-config-table td:first-child,
-#save-settings-config-table th:first-child {
+.form-table td:first-child > *,
+.form-table th:first-child > * {
    float: right;
+    white-space: nowrap;
 }

-#save-settings-config-table td:last-child,
-#save-settings-config-table th:last-child {
+.form-table td:last-child > *,
+.form-table th:last-child > * {
    float: left;
 }

-#save-settings-config-table td small {
+.form-table small {
    color: rgb(153, 153, 153);
 }

-#save-settings-config-close-btn {
-    float: right;
-    cursor: pointer;
-    padding: 10px;
-    transform: translate(50%, -50%) scaleX(130%);
+#system-settings .form-table td {
+    height: 24px;
 }

-#reset-image-settings {
-    cursor: pointer;
-    float: right;
-    padding: 8px;
-    opacity: 1;
-    transition: opacity 0.5;
+#system-settings .form-table td:last-child div {
+    display: flex;
+    align-items: center;
+}
+#system-settings .form-table td:last-child div > :not([type="checkbox"]):first-child {
+    margin-left: 3px;
 }

-.collapsible:not(.active) #reset-image-settings {
-    display: none;
-}
-
-#reset-image-settings.hidden {
-    opacity: 0;
-    pointer-events: none;
-}
+#system-settings .form-table td:last-child div small {
+    padding-left: 5px;
+    text-align: left;
+}
--- a/ui/media/css/main.css
+++ b/ui/media/css/main.css
@ -8,6 +8,7 @@ html {
 }

 body {
+    margin: 0;
    font-size: 11pt;
    background-color: var(--background-color1);
    color: var(--text-color);
@ -26,9 +27,10 @@ label {
    height: 65pt;
    font-size: 13px;
    margin-bottom: 6px;
+    margin-top: 5px;
+    display: block;
 }
 .image_preview_container {
-    /* display: none; */
    margin-top: 10pt;
 }
 .image_clear_btn {
@ -64,17 +66,17 @@ label {
    font-size: small;
    padding-bottom: 3pt;
 }
-#progressBar {
-    font-size: small;
-}
 #footer {
    font-size: small;
-    padding-left: 10pt;
+    padding: 10pt;
    background: none;
 }
 #footer-legal {
    font-size: 8pt;
 }
+#footer-spacer {
+    flex: 0.7
+}
 .imgSeedLabel {
    font-size: 0.8em;
    background-color: var(--background-color2);
@ -107,33 +109,42 @@ label {
    margin-bottom: 7px;
 }
 #container {
-    width: 95%;
-    margin-left: auto;
-    margin-right: auto;
-}
-@media screen and (max-width: 1800px) {
-    #container {
-        width: 100%;
-    }
+    min-height: 100vh;
+    width: 100%;
+    margin: 0px;
+    display: flex;
+    flex-direction: column;
 }
 #logo small {
    font-size: 11pt;
 }
 #editor {
-    padding: 5px;
+    background: var(--background-color1);
+    padding: 16px;
+    display: flex;
+    flex-direction: column;
+    flex: 0 0 370pt;
 }
 #editor label {
    font-weight: normal;
 }
+#editor h4 {
+    margin: 0px;
+    white-space: nowrap;
+}
+#editor .collapsible-content {
+    width: 100%;
+}
 .settings-box label small {
    color: rgb(153, 153, 153);
    margin-right: 10px;
 }
 #preview {
-    padding: 5px;
+    padding: 8px;
+    background: var(--background-color1);
 }
-#editor-inputs {
-    margin-bottom: 20px;
+#preview .collapsible-content {
+    padding: 0px 15px;
 }
 #editor-inputs-prompt {
    flex: 1;
@ -151,7 +162,7 @@ label {
 #makeImage {
    flex: 0 0 70px;
    background: var(--accent-color);
-    border: var(--make-image-border);
+    border: var(--primary-button-border);
    color: rgb(255, 221, 255);
    width: 100%;
    height: 30pt;
@ -168,6 +179,7 @@ label {
    height: 30pt;
    border-radius: 6px;
    display: none;
+    margin-top: 2pt;
 }
 #stopImage:hover {
    background: rgb(177, 27, 0);
@ -176,12 +188,6 @@ label {
    display: flex;
    width: 100%;
 }
-.col-50 {
-    flex: 50%;
-}
-.col-fixed-10 {
-    flex: 0 0 350pt;
-}
 .col-free {
    flex: 1;
 }
@ -189,7 +195,7 @@ label {
    cursor: pointer;
 }
 .collapsible-content {
-    display: none;
+    display: block;
    padding-left: 15px;
 }
 .collapsible-content h5 {
@ -201,50 +207,40 @@ label {
    color: white;
    padding-right: 5px;
 }
-.panel-box {
-    background: var(--background-color2);
-    border: 1px solid var(--background-color3);
-    border-radius: 7px;
-    padding: 5px;
-    margin-bottom: 15px;
-    box-shadow: 0 4px 8px 0 rgba(0, 0, 0, 0.15), 0 6px 20px 0 rgba(0, 0, 0, 0.15);
+.collapsible:not(.active) ~ .collapsible-content  {
+    display: none !important;
 }
-.panel-box h4 {
-    margin: 0;
-    padding: 2px 0;
+#editor-modifiers {
+    max-width: 600px;
+    overflow-y: auto;
+    overflow-x: hidden;
 }
 #editor-modifiers .editor-modifiers-leaf {
    padding-top: 10pt;
    padding-bottom: 10pt;
 }
-#preview {
-    margin-left: 10pt;
-}
 img {
    box-shadow: 0 4px 8px 0 rgba(0, 0, 0, 0.15), 0 6px 20px 0 rgba(0, 0, 0, 0.15);
 }
 .line-separator {
-    background: rgb(56, 56, 56);
+    background: var(--background-color3);
    height: 1pt;
-    margin: 15pt 0;
+    margin: 16px 0px;
 }
 #editor-inputs-tags-container {
    margin-top: 5pt;
    display: none;
 }
 #server-status {
-    display: inline;
-    float: right;
-    transform: translateY(-5pt);
+    position: absolute;
+    right: 16px;
+    top: 50%;
+    transform: translateY(-50%);
+    text-align: right;
 }
 #server-status-color {
-    /* width: 8pt;
-    height: 8pt;
-    border-radius: 4pt; */
    font-size: 14pt;
    color: rgb(200, 139, 0);
-    /* background-color: rgb(197, 1, 1); */
-    /* transform: translateY(15%); */
    display: inline;
 }
 #server-status-msg {
@ -288,16 +284,19 @@ img {
 }

 #top-nav {
-    padding-top: 3pt;
-    padding-bottom: 15pt;
+    position: relative;
+    background: var(--background-color4);
+    display: flex;
 }
-#top-nav .icon {
+.tab .icon {
    padding-right: 4pt;
    font-size: 14pt;
    transform: translateY(1pt);
 }
 #logo {
    display: inline;
+    padding: 12px;
+    white-space: nowrap;
 }
 #logo h1 {
    display: inline;
@ -311,6 +310,8 @@ img {
    float: left;
    display: inline;
    padding-left: 20pt;
+}
+#top-nav-items > li:first-child {
    cursor: default;
 }
 #initial-text {
@ -324,26 +325,12 @@ img {
 .pl-5 {
    padding-left: 5pt;
 }
-#system-settings {
-    width: 360pt;
-    transform: translateX(-100%) translateX(70pt);
-
-    padding-top: 10pt;
-    padding-bottom: 10pt;
-}
-#system-settings ul {
-    margin: 0;
-    padding: 0;
-}
-#system-settings li {
-    padding-left: 5pt;
-}
 #community-links {
+    display: inline-block;
    list-style-type: none;
-    margin: 0;
-    padding: 12pt;
-    padding-bottom: 0pt;
-    transform: translateX(-15%);
+    text-align: left;
+    margin: auto;
+    padding: 0px;
 }
 #community-links li {
    padding-bottom: 12pt;
@ -357,6 +344,35 @@ img {
    color: var(--text-color);
    text-decoration: none;
 }
+.float-child h1 {
+    border-bottom: var(--button-border);
+}
+#help-links {
+    display: inline-block;
+    list-style-type: none;
+    text-align: left;
+    margin: auto;
+    padding: 0px;
+}
+#help-links li {
+    padding-bottom: 12pt;
+    display: block;
+    font-size: 10pt;
+}
+#help-links li .fa-fw {
+    padding-right: 2pt;
+}
+#help-links li a {
+    color: var(--text-color);
+    text-decoration: none;
+}
+#help-links li ul {
+    padding-inline-start: 10px;
+    margin-top: 8px;
+}
+.help-section {
+    font-size: 130%;
+}
 .dropdown {
    overflow: hidden;
 }
@ -383,6 +399,9 @@ img {
    border-radius: 5pt;
    box-shadow: 0 20px 28px 0 rgba(0, 0, 0, 0.15), 0 6px 20px 0 rgba(0, 0, 0, 0.15);
 }
+.imageTaskContainer > div > .collapsible-handle {
+    display: none;
+}
 .taskStatusLabel {
    float: left;
    font-size: 8pt;
@ -402,6 +421,12 @@ img {
    border: 1px solid rgb(107, 75, 0);
    color:rgb(255, 242, 211)
 }
+.primaryButton {
+    flex: 0 0 70px;
+    background: var(--accent-color);
+    border: var(--primary-button-border);
+    color: rgb(255, 221, 255);
+}
 .secondaryButton {
    background: rgb(132, 8, 0);
    border: 1px solid rgb(122, 29, 0);
@ -433,6 +458,8 @@ img {
 #init_image_preview {
    max-width: 150px;
    max-height: 150px;
+    height: 100%;
+    width: 100%;
 	object-fit: contain;
    border-radius: 6px;
    transition: all 1s ease-in-out;
@ -441,6 +468,7 @@ img {
 #init_image_preview:hover {
    max-width: 500px;
    max-height: 1000px;
+
    transition: all 1s 0.5s ease-in-out;
 }

@ -462,6 +490,20 @@ img {
    border-radius: 6px 0px;
 }

+#editor-settings-entries {
+    display: flex;
+    flex-direction: column;
+}
+
+#editor-settings-entries > div {
+    margin-top: 15px;
+}
+
+#editor-settings-entries ul {
+    margin: 0px;
+    padding: 0px;
+}
+
 #editor-settings-entries table td {
    padding: 0px;
    line-height: 28px;
@ -477,6 +519,7 @@ img {
    width: 100%;
 }

+/* INPUTS STYLING */
 button,
 input[type="file"],
 input[type="checkbox"],
@ -536,13 +579,16 @@ input::file-selector-button {
    height: 19px;
 }

+/* MOBILE SUPPORT */
@media screen and (max-width: 700px) {
+    #top-nav {
+        flex-direction: column;
+    }
    body {
        margin: 0px;
    }
    #container {
        margin: 0px;
-        padding: 10px
    }
    .flex-container {
        flex-direction: column;
@ -571,21 +617,98 @@ input::file-selector-button {
        left: 0px;
        right: 0px;
    }
+    #editor {
+        padding: 16px 8px;
+    }
+    .tab-content-inner {
+        margin: 0px;
+    }
+    .tab {
+        font-size: 0;
+    }
+    .tab .icon {
+        padding-right: 0px;
+    }
+    #server-status {
+        display: none;
+    }
+    .popup > div {
+        padding-left: 5px !important;
+        padding-right: 5px !important;
+    }
+    .popup > div input, .popup > div select {
+        max-width: 40vw;
+    }
+    .popup .close-button {
+        padding: 0px !important;
+        margin: 24px !important;
+    }
+    .simple-tooltip.right {
+        right: initial;
+        left: 0px;
+        top: 50%;
+        transform: translate(calc(-100% + 15%), -50%);
+    }
+    :hover > .simple-tooltip.right {
+        transform: translate(100%, -50%);
+    }
+}
+
+@media (min-width: 700px) {
+    /* #editor {
+        max-width: 480px;
+    } */
+    .float-container {
+        padding: 20px;
+    }
+    .float-child {
+        width: 50%;
+        float: left;
+        padding: 20px;
+    }
+}
+
+.help-btn {
+    position: relative;
 }

 #promptsFromFileBtn {
    font-size: 9pt;
 }

-#reset-image-settings {
+.section-button {
    position: relative;
    transform: translateY(-13%);
 }
+.collapsible:not(.active) #copy-image-settings {
+    display: none;
+}

+.section-button {
+    cursor: pointer;
+    float: right;
+    padding: 8px;
+    opacity: 1;
+    transition: opacity 0.5;
+}
+
+.section-button {
+    cursor: pointer;
+    float: right;
+    padding: 8px;
+    opacity: 1;
+    transition: opacity 0.5;
+}
+
+.collapsible:not(.active) .section-button {
+    display: none;
+}
+
+/* SIMPLE TOOTIP */
 .simple-tooltip {
 	border-radius: 3px;
 	font-weight: bold;
-	font-size: 16px;
+	font-size: 12px;
    background-color: var(--background-color3);

 	visibility: hidden;
@ -604,8 +727,6 @@ input::file-selector-button {
        visibility: visible;
    }
 }
-
-/* position specific */
 .simple-tooltip.right {
    right: 0px;
    top: 50%;
@ -641,3 +762,154 @@ input::file-selector-button {
 :hover > .simple-tooltip.bottom {
    transform: translate(-50%, 100%);
 }
+
+/* PROGRESS BAR */
+.progress-bar {
+    background: var(--background-color3);
+    border-radius: 4px;
+    border: 2px solid var(--background-color3);
+    height: 16px;
+    position: relative;
+    transition: 0.25s 1s border, 0.25s 1s height;
+}
+.progress-bar > div {
+    background: var(--accent-color);
+    border-radius: 4px;
+    position: absolute;
+    left: 0;
+    top: 0;
+    bottom: 0;
+    width: 0%;
+    transition: width 1s ease-in-out;
+}
+.progress-bar.active {
+    background: repeating-linear-gradient(-65deg, 
+        var(--background-color2), 
+        var(--background-color2) 4px, 
+        var(--background-color3) 5px, 
+        var(--background-color3) 9px, 
+        var(--background-color2) 10px);
+    background-size: 200% auto;
+    background-position: 0 100%;
+    animation: progress-anim 2s infinite;
+    animation-fill-mode: forwards;
+    animation-timing-function: linear;
+}
+
+@keyframes progress-anim { 
+    0%   { background-position: -55px 0; }
+    100% { background-position: 0 0; }
+}
+
+/* POPUPS */
+.popup:not(.active) {
+    visibility: hidden;
+    opacity: 0;
+}
+
+.popup {
+    position: absolute;
+    background: rgba(32, 33, 36, 50%);
+    top: 0px;
+    left: 0px;
+    right: 0px;
+    bottom: 0px;
+    z-index: 1000;
+    opacity: 1;
+    transition: 0s visibility, 0.3s opacity;
+}
+
+@media only screen and (min-height: 1050px) {
+    .popup {
+        position: fixed;
+    }
+}
+
+.popup > div {
+    position: relative;
+    background: var(--background-color2);
+    border: solid 1px var(--background-color3);
+    max-width: 700px;
+    margin: auto;
+    margin-top: 50px;
+    border-radius: 6px;
+    padding: 30px;
+    text-align: center;
+    box-shadow: 0px 0px 30px black;
+}
+
+.popup .close-button {
+    position: absolute;
+    right: 0px;
+    top: 0px;
+    transform: scale(150%);
+    cursor: pointer;
+    padding: 24px;
+}
+
+/* TABS */
+#tab-container {
+    display: flex;
+    align-items: flex-end;
+}
+
+.tab {
+    padding: 8px 16px;
+    border-radius: 4px 4px 0px 0px;
+    margin-left: 8px;
+    cursor: pointer;
+    background: var(--background-color1);
+    opacity: 50%;
+    transition: opacity 0.25s;
+}
+
+.tab:hover {
+    opacity: 75%;
+}
+
+.tab.active {
+    opacity: 100%;
+}
+
+.tab-content:not(.active) {
+    display: none;
+}
+
+#tab-content-wrapper {
+    border-top: 8px solid var(--background-color1);
+}
+
+.tab-content-inner {
+    margin: auto;
+    max-width: 600px;
+    text-align: center;
+    padding: 20px 10px;
+}
+
+.panel-box {
+    background: var(--background-color2);
+    border: 1px solid var(--background-color3);
+    border-radius: 7px;
+    padding: 7px;
+    margin-bottom: 15px;
+    box-shadow: 0 4px 8px 0 rgba(0, 0, 0, 0.15), 0 6px 20px 0 rgba(0, 0, 0, 0.15);
+}
+
+i.active {
+    background: var(--accent-color);
+}
+#system-info {
+    max-width: 800px;
+    font-size: 10pt;
+}
+#system-info .value {
+    text-align: left;
+    padding-left: 10pt;
+}
+#system-info label {
+    float: right;
+    font-weight: bold;
+}
+#save-system-settings-btn {
+    padding: 4pt 8pt;
+}
--- a/ui/media/css/modifier-thumbnails.css
+++ b/ui/media/css/modifier-thumbnails.css
@ -217,32 +217,6 @@
 #modifier-settings-btn {
    float: right;
 }
-
-#modifier-settings-config {
-    position: fixed;
-    background: rgba(32, 33, 36, 50%);
-    top: 0px;
-    left: 0px;
-    width: 100vw;
-    height: 100vh;
-    z-index: 1000;
-}
-
-#modifier-settings-config > div {
-    background: var(--background-color2);
-    max-width: 600px;
-    margin: auto;
-    margin-top: 100px;
-    border-radius: 6px;
-    padding: 30px;
-    text-align: center;
-}
-#modifier-settings-config-close-btn {
-    float: right;
-    cursor: pointer;
-    padding: 10px;
-    transform: translate(50%, -50%) scaleX(130%);
-}
 #modifier-settings-config textarea {
    width: 90%;
    height: 150px;
--- a/ui/media/css/themes.css
+++ b/ui/media/css/themes.css
@ -23,12 +23,12 @@
    --input-border-size: 1px; 
    --accent-color: hsl(var(--accent-hue), 100%, var(--accent-lightness));
    --accent-color-hover: hsl(var(--accent-hue), 100%, var(--accent-lightness-hover));
-    --make-image-border: 2px solid hsl(var(--accent-hue), 100%, calc(var(--accent-lightness) - 21%));
+    --primary-button-border: none;
 }

 .theme-light {
    --background-color1: white; 
-    --background-color2: #dddddd;
+    --background-color2: #ececec;
    --background-color3: #e7e9eb;
    --background-color4: #cccccc;

@ -47,7 +47,7 @@
    
    --accent-hue: 235;
    --accent-lightness: 65%;
-    --make-image-border: none;
+    --primary-button-border: none;
    
    --button-color: var(--accent-color);
    --button-border: none;
@ -61,7 +61,7 @@
 .theme-cool-blue {
    --main-hue: 222;
 	--main-saturation: 18%;
-	--value-base: 19%;
+	--value-base: 18%;
 	--value-step: 3%;
 	--background-color1: hsl(var(--main-hue), var(--main-saturation), var(--value-base));
 	--background-color2: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) - (1 * var(--value-step))));
@ -69,7 +69,7 @@
 	--background-color4: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) - (3 * var(--value-step))));
    
    --accent-hue: 212;
-    --make-image-border: none;
+    --primary-button-border: none;

    --button-color: var(--accent-color);
    --button-border: none;
@ -91,7 +91,7 @@
 	--background-color3: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) - (2 * var(--value-step))));
 	--background-color4: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) - (3 * var(--value-step))));
    
-    --make-image-border: none;
+    --primary-button-border: none;

    --button-color: var(--accent-color);
    --button-border: none;
@ -110,9 +110,9 @@
 	--background-color1: hsl(var(--main-hue), var(--main-saturation), var(--value-base));
 	--background-color2: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) + (1 * var(--value-step))));
 	--background-color3: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) + (2 * var(--value-step))));
-	--background-color4: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) + (3 * var(--value-step))));
+	--background-color4: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) + (1.4 * var(--value-step))));
    
-    --make-image-border: none;
+    --primary-button-border: none;

    --button-color: var(--accent-color);
    --button-border: none;
@ -134,7 +134,7 @@
 	--background-color4: hsl(var(--main-hue), var(--main-saturation), calc(var(--value-base) - (3 * var(--value-step))));
    
    --accent-hue: 212;
-    --make-image-border: none;
+    --primary-button-border: none;

    --button-color: var(--accent-color);
    --button-border: none;
--- a/ui/media/js/auto-save.js
+++ b/ui/media/js/auto-save.js
@ -13,6 +13,7 @@ const SETTINGS_IDS_LIST = [
    "num_outputs_total",
    "num_outputs_parallel",
    "stable_diffusion_model",
+    "vae_model",
    "sampler",
    "width",
    "height",
@ -33,7 +34,6 @@ const SETTINGS_IDS_LIST = [
    "diskPath",
    "sound_toggle",
    "turbo",
-    "use_cpu",
    "use_full_precision",
    "auto_save_settings"
 ]
@ -52,7 +52,9 @@ const SETTINGS_SECTIONS = [ // gets the "keys" property filled in with an ordere
 async function initSettings() {
    SETTINGS_IDS_LIST.forEach(id => {
        var element = document.getElementById(id)
-        var label = document.querySelector(`label[for='${element.id}']`)
+        if (!element) {
+            console.error(`Missing settings element ${id}`)
+        }
        SETTINGS[id] = {
            key: id,
            element: element,
@ -68,7 +70,8 @@ async function initSettings() {
    SETTINGS_SECTIONS.forEach(section => {
        var name = section.name
        var element = document.getElementById(section.id)
-        var children = Array.from(element.querySelectorAll(unsorted_settings_ids.map(id => `#${id}`).join(",")))
+        var unsorted_ids = unsorted_settings_ids.map(id => `#${id}`).join(",")
+        var children = unsorted_ids == "" ? [] : Array.from(element.querySelectorAll(unsorted_ids));
        section.keys = []
        children.forEach(e => {
            section.keys.push(e.id)
@ -126,10 +129,11 @@ function loadSettings() {
            return
        }
        CURRENTLY_LOADING_SETTINGS = true
-        saved_settings.map(saved_setting => {
+        saved_settings.forEach(saved_setting => {
            var setting = SETTINGS[saved_setting.key]
-            if (setting === undefined) {
-                return
+            if (!setting) {
+                console.warn(`Attempted to load setting ${saved_setting.key}, but no setting found`);
+                return null;
            }
            setting.ignore = saved_setting.ignore
            if (!setting.ignore) {
@ -211,20 +215,22 @@ function fillSaveSettingsConfigTable() {
    })
 }

-document.getElementById("save-settings-config-close-btn").addEventListener('click', () => {
-    saveSettingsConfigOverlay.style.display = 'none'
+// configureSettingsSaveBtn
+
+
+
+
+var autoSaveSettings = document.getElementById("auto_save_settings")
+var configSettingsButton = document.createElement("button")
+configSettingsButton.textContent = "Configure"
+configSettingsButton.style.margin = "0px 5px"
+autoSaveSettings.insertAdjacentElement("afterend", configSettingsButton)
+autoSaveSettings.addEventListener("change", () => {
+    configSettingsButton.style.display = autoSaveSettings.checked ? "block" : "none"
 })
-document.getElementById("configureSettingsSaveBtn").addEventListener('click', () => {
+configSettingsButton.addEventListener('click', () => {
    fillSaveSettingsConfigTable()
-    saveSettingsConfigOverlay.style.display = 'block'
-})
-saveSettingsConfigOverlay.addEventListener('click', (event) => {
-    if (event.target.id == saveSettingsConfigOverlay.id) {
-        saveSettingsConfigOverlay.style.display = 'none'
-    }
-})
-document.getElementById("save-settings-config-close-btn").addEventListener('click', () => {
-    saveSettingsConfigOverlay.style.display = 'none'
+    saveSettingsConfigOverlay.classList.add("active")
 })
 resetImageSettingsButton.addEventListener('click', event => {
    loadDefaultSettingsSection("editor-settings");
@ -276,9 +282,11 @@ function tryLoadOldSettings() {
    Object.keys(individual_settings_map).forEach(localStorageKey => {
        var localStorageValue = localStorage.getItem(localStorageKey);
        if (localStorageValue !== null) {
-            var setting = SETTINGS[individual_settings_map[localStorageKey]]
-            if (setting == null || setting == undefined) {
-                return
+            let key = individual_settings_map[localStorageKey]
+            var setting = SETTINGS[key]
+            if (!setting) {
+                console.warn(`Attempted to map old setting ${key}, but no setting found`);
+                return null;
            }
            if (setting.element.type == "checkbox" && (typeof localStorageValue === "string" || localStorageValue instanceof String)) {
                localStorageValue = localStorageValue == "true"
--- a/ui/media/js/dnd.js
+++ b/ui/media/js/dnd.js
@ -0,0 +1,469 @@
+"use strict" // Opt in to a restricted variant of JavaScript
+
+const EXT_REGEX = /(?:\.([^.]+))?$/
+const TEXT_EXTENSIONS = ['txt', 'json']
+const IMAGE_EXTENSIONS = ['jpg', 'jpeg', 'png', 'bmp', 'tiff', 'tif', 'tga']
+
+function parseBoolean(stringValue) {
+    if (typeof stringValue === 'boolean') {
+        return stringValue
+    }
+    if (typeof stringValue === 'number') {
+        return stringValue !== 0
+    }
+    if (typeof stringValue !== 'string') {
+        return false
+    }
+    switch(stringValue?.toLowerCase()?.trim()) {
+        case "true":
+        case "yes":
+        case "on":
+        case "1":
+          return true;
+
+        case "false":
+        case "no":
+        case "off":
+        case "0":
+        case null:
+        case undefined:
+          return false;
+    }
+    try {
+        return Boolean(JSON.parse(stringValue));
+    } catch {
+        return Boolean(stringValue)
+    }
+}
+
+const TASK_MAPPING = {
+    prompt: { name: 'Prompt',
+        setUI: (prompt) => {
+            promptField.value = prompt
+        },
+        readUI: () => promptField.value,
+        parse: (val) => val
+    },
+    negative_prompt: { name: 'Negative Prompt',
+        setUI: (negative_prompt) => {
+            negativePromptField.value = negative_prompt
+        },
+        readUI: () => negativePromptField.value,
+        parse: (val) => val
+    },
+    width: { name: 'Width',
+        setUI: (width) => {
+            const oldVal = widthField.value
+            widthField.value = width
+            if (!widthField.value) {
+                widthField.value = oldVal
+            }
+        },
+        readUI: () => parseInt(widthField.value),
+        parse: (val) => parseInt(val)
+    },
+    height: { name: 'Height',
+        setUI: (height) => {
+            const oldVal = heightField.value
+            heightField.value = height
+            if (!heightField.value) {
+                heightField.value = oldVal
+            }
+        },
+        readUI: () => parseInt(heightField.value),
+        parse: (val) => parseInt(val)
+    },
+    seed: { name: 'Seed',
+        setUI: (seed) => {
+            if (!seed) {
+                randomSeedField.checked = true
+                seedField.disabled = true
+                return
+            }
+            randomSeedField.checked = false
+            seedField.disabled = false
+            seedField.value = seed
+        },
+        readUI: () => (randomSeedField.checked ? Math.floor(Math.random() * 10000000) : parseInt(seedField.value)),
+        parse: (val) => parseInt(val)
+    },
+    num_inference_steps: { name: 'Steps',
+        setUI: (num_inference_steps) => {
+            numInferenceStepsField.value = num_inference_steps
+        },
+        readUI: () => parseInt(numInferenceStepsField.value),
+        parse: (val) => parseInt(val)
+    },
+    guidance_scale: { name: 'Guidance Scale',
+        setUI: (guidance_scale) => {
+            guidanceScaleField.value = guidance_scale
+            updateGuidanceScaleSlider()
+        },
+        readUI: () => parseFloat(guidanceScaleField.value),
+        parse: (val) => parseFloat(val)
+    },
+    prompt_strength: { name: 'Prompt Strength',
+        setUI: (prompt_strength) => {
+            promptStrengthField.value = prompt_strength
+            updatePromptStrengthSlider()
+        },
+        readUI: () => parseFloat(promptStrengthField.value),
+        parse: (val) => parseFloat(val)
+    },
+
+    init_image:  { name: 'Initial Image',
+        setUI: (init_image) => {
+            initImagePreview.src = init_image
+        },
+        readUI: () => initImagePreview.src,
+        parse: (val) => val
+    },
+    mask:  { name: 'Mask',
+        setUI: (mask) => {
+            inpaintingEditor.setImg(mask)
+            maskSetting.checked = Boolean(mask)
+        },
+        readUI: () => (maskSetting.checked ? inpaintingEditor.getImg() : undefined),
+        parse: (val) => val
+    },
+
+    use_face_correction: { name: 'Use Face Correction',
+        setUI: (use_face_correction) => {
+            useFaceCorrectionField.checked = parseBoolean(use_face_correction)
+        },
+        readUI: () => useFaceCorrectionField.checked,
+        parse: (val) => parseBoolean(val)
+    },
+    use_upscale: { name: 'Use Upscaling',
+        setUI: (use_upscale) => {
+            const oldVal = upscaleModelField.value
+            upscaleModelField.value = use_upscale
+            if (upscaleModelField.value) { // Is a valid value for the field.
+                useUpscalingField.checked = true
+                upscaleModelField.disabled = false
+            } else { // Not a valid value, restore the old value and disable the filter.
+                upscaleModelField.disabled = true
+                upscaleModelField.value = oldVal
+                useUpscalingField.checked = false
+            }
+        },
+        readUI: () => (useUpscalingField.checked ? upscaleModelField.value : undefined),
+        parse: (val) => val
+    },
+    sampler: { name: 'Sampler',
+        setUI: (sampler) => {
+            samplerField.value = sampler
+        },
+        readUI: () => samplerField.value,
+        parse: (val) => val
+    },
+    use_stable_diffusion_model: { name: 'Stable Diffusion model',
+        setUI: (use_stable_diffusion_model) => {
+            const oldVal = stableDiffusionModelField.value
+
+            let pathIdx = use_stable_diffusion_model.lastIndexOf('/') // Linux, Mac paths
+            if (pathIdx < 0) {
+                pathIdx = use_stable_diffusion_model.lastIndexOf('\\') // Windows paths.
+            }
+            if (pathIdx >= 0) {
+                use_stable_diffusion_model = use_stable_diffusion_model.slice(pathIdx + 1)
+            }
+            const modelExt = '.ckpt'
+            if (use_stable_diffusion_model.endsWith(modelExt)) {
+                use_stable_diffusion_model = use_stable_diffusion_model.slice(0, use_stable_diffusion_model.length - modelExt.length)
+            }
+
+            stableDiffusionModelField.value = use_stable_diffusion_model
+
+            if (!stableDiffusionModelField.value) {
+                stableDiffusionModelField.value = oldVal
+            }
+        },
+        readUI: () => stableDiffusionModelField.value,
+        parse: (val) => val
+    },
+
+    numOutputsParallel: { name: 'Parallel Images',
+        setUI: (numOutputsParallel) => {
+            numOutputsParallelField.value = numOutputsParallel
+        },
+        readUI: () => parseInt(numOutputsParallelField.value),
+        parse: (val) => val
+    },
+
+    use_cpu: { name: 'Use CPU',
+        setUI: (use_cpu) => {
+            useCPUField.checked = use_cpu
+        },
+        readUI: () => useCPUField.checked,
+        parse: (val) => val
+    },
+    turbo: { name: 'Turbo',
+        setUI: (turbo) => {
+            turboField.checked = turbo
+        },
+        readUI: () => turboField.checked,
+        parse: (val) => Boolean(val)
+    },
+    use_full_precision: { name: 'Use Full Precision',
+        setUI: (use_full_precision) => {
+            useFullPrecisionField.checked = use_full_precision
+        },
+        readUI: () => useFullPrecisionField.checked,
+        parse: (val) => Boolean(val)
+    },
+
+    stream_image_progress: { name: 'Stream Image Progress',
+        setUI: (stream_image_progress) => {
+            streamImageProgressField.checked = (parseInt(numOutputsTotalField.value) > 50 ? false : stream_image_progress)
+        },
+        readUI: () => streamImageProgressField.checked,
+        parse: (val) => Boolean(val)
+    },
+    show_only_filtered_image: { name: 'Show only the corrected/upscaled image',
+        setUI: (show_only_filtered_image) => {
+            showOnlyFilteredImageField.checked = show_only_filtered_image
+        },
+        readUI: () => showOnlyFilteredImageField.checked,
+        parse: (val) => Boolean(val)
+    },
+    output_format: { name: 'Output Format',
+        setUI: (output_format) => {
+            outputFormatField.value = output_format
+        },
+        readUI: () => outputFormatField.value,
+        parse: (val) => val
+    },
+    save_to_disk_path: { name: 'Save to disk path',
+        setUI: (save_to_disk_path) => {
+            saveToDiskField.checked = Boolean(save_to_disk_path)
+            diskPathField.value = save_to_disk_path
+        },
+        readUI: () => diskPathField.value,
+        parse: (val) => val
+    }
+}
+function restoreTaskToUI(task) {
+    if ('numOutputsTotal' in task) {
+        numOutputsTotalField.value = task.numOutputsTotal
+    }
+    if ('seed' in task) {
+        randomSeedField.checked = false
+        seedField.value = task.seed
+    }
+    if (!('reqBody' in task)) {
+        return
+    }
+    for (const key in TASK_MAPPING) {
+        if (key in task.reqBody) {
+            TASK_MAPPING[key].setUI(task.reqBody[key])
+        }
+    }
+}
+function readUI() {
+    const reqBody = {}
+    for (const key in TASK_MAPPING) {
+        reqBody[key] = TASK_MAPPING[key].readUI()
+    }
+    return {
+        'numOutputsTotal': parseInt(numOutputsTotalField.value),
+        'seed': TASK_MAPPING['seed'].readUI(),
+        'reqBody': reqBody
+    }
+}
+
+const TASK_TEXT_MAPPING = {
+    width: 'Width',
+    height: 'Height',
+    seed: 'Seed',
+    num_inference_steps: 'Steps',
+    guidance_scale: 'Guidance Scale',
+    prompt_strength: 'Prompt Strength',
+    use_face_correction: 'Use Face Correction',
+    use_upscale: 'Use Upscaling',
+    sampler: 'Sampler',
+    negative_prompt: 'Negative Prompt',
+    use_stable_diffusion_model: 'Stable Diffusion model'
+}
+const afterPromptRe = /^\s*Width\s*:\s*\d+\s*(?:\r\n|\r|\n)+\s*Height\s*:\s*\d+\s*(\r\n|\r|\n)+Seed\s*:\s*\d+\s*$/igm
+function parseTaskFromText(str) {
+    const taskReqBody = {}
+
+    // Prompt
+    afterPromptRe.lastIndex = 0
+    const match = afterPromptRe.exec(str)
+    if (match) {
+        let prompt = str.slice(0, match.index)
+        str = str.slice(prompt.length)
+        taskReqBody.prompt = prompt.trim()
+        console.log('Prompt:', taskReqBody.prompt)
+    }
+    for (const key in TASK_TEXT_MAPPING) {
+        const name = TASK_TEXT_MAPPING[key];
+        let val = undefined
+
+        const reName = new RegExp(`${name}\\ *:\\ *(.*)(?:\\r\\n|\\r|\\n)*`, 'igm')
+        const match = reName.exec(str);
+        if (match) {
+            str = str.slice(0, match.index) + str.slice(match.index + match[0].length)
+            val = match[1]
+        }
+        if (val !== undefined) {
+            taskReqBody[key] = TASK_MAPPING[key].parse(val.trim())
+            console.log(TASK_MAPPING[key].name + ':', taskReqBody[key])
+            if (!str) {
+                break
+            }
+        }
+    }
+    if (Object.keys(taskReqBody).length <= 0) {
+        return undefined
+    }
+    const task = { reqBody: taskReqBody }
+    if ('seed' in taskReqBody) {
+        task.seed = taskReqBody.seed
+    }
+    return task
+}
+
+async function readFile(file, i) {
+    const fileContent = (await file.text()).trim()
+
+    // JSON File.
+    if (fileContent.startsWith('{') && fileContent.endsWith('}')) {
+        try {
+            const task = JSON.parse(fileContent)
+            restoreTaskToUI(task)
+        } catch (e) {
+            console.warn(`file[${i}]:${file.name} - File couldn't be parsed.`, e)
+        }
+        return
+    }
+
+    // Normal txt file.
+    const task = parseTaskFromText(fileContent)
+    if (task) {
+        restoreTaskToUI(task)
+    } else {
+        console.warn(`file[${i}]:${file.name} - File couldn't be parsed.`)
+    }
+}
+
+function dropHandler(ev) {
+    console.log('Content dropped...')
+    let items = []
+
+    if (ev?.dataTransfer?.items) { // Use DataTransferItemList interface
+        items = Array.from(ev.dataTransfer.items)
+        items = items.filter(item => item.kind === 'file')
+        items = items.map(item => item.getAsFile())
+    } else if (ev?.dataTransfer?.files) { // Use DataTransfer interface
+        items = Array.from(ev.dataTransfer.files)
+    }
+
+    items.forEach(item => {item.file_ext = EXT_REGEX.exec(item.name.toLowerCase())[1]})
+
+    let text_items = items.filter(item => TEXT_EXTENSIONS.includes(item.file_ext))
+    let image_items = items.filter(item => IMAGE_EXTENSIONS.includes(item.file_ext))
+
+    if (image_items.length > 0 && ev.target == initImageSelector) {
+        return // let the event bubble up, so that the Init Image filepicker can receive this
+    }
+
+    ev.preventDefault() // Prevent default behavior (Prevent file/content from being opened)
+    text_items.forEach(readFile)
+}
+function dragOverHandler(ev) {
+    console.log('Content in drop zone')
+
+    // Prevent default behavior (Prevent file/content from being opened)
+    ev.preventDefault()
+
+    ev.dataTransfer.dropEffect = "copy"
+
+    let img = new Image()
+    img.src = location.host + '/media/images/favicon-32x32.png'
+    ev.dataTransfer.setDragImage(img, 16, 16)
+}
+
+document.addEventListener("drop", dropHandler)
+document.addEventListener("dragover", dragOverHandler)
+
+const TASK_REQ_NO_EXPORT = [
+    "use_cpu",
+    "turbo",
+    "use_full_precision",
+    "save_to_disk_path"
+]
+
+// Retrieve clipboard content and try to parse it 
+async function pasteFromClipboard() {
+    //const text = await navigator.clipboard.readText()
+    let text = await navigator.clipboard.readText();
+    text=text.trim();
+    if (text.startsWith('{') && text.endsWith('}')) {
+        try {
+            const task = JSON.parse(text)
+            restoreTaskToUI(task)
+        } catch (e) {
+            console.warn(`Clipboard JSON couldn't be parsed.`, e)
+        }
+        return
+    }
+    // Normal txt file.
+    const task = parseTaskFromText(text)
+    if (task) {
+        restoreTaskToUI(task)
+    } else {
+        console.warn(`Clipboard content - File couldn't be parsed.`)
+    }
+}
+
+// Adds a copy and a paste icon if the browser grants permission to write to clipboard.
+function checkWriteToClipboardPermission (result) {
+    if (result.state == "granted" || result.state == "prompt") {
+        const resetSettings = document.getElementById('reset-image-settings')
+
+        // COPY ICON
+        const copyIcon = document.createElement('i')
+        copyIcon.className = 'fa-solid fa-clipboard section-button'
+        copyIcon.innerHTML = `<span class="simple-tooltip right">Copy Image Settings</span>`
+        copyIcon.addEventListener('click', (event) => {
+            event.stopPropagation()
+	    // Add css class 'active'
+	    copyIcon.classList.add('active')
+	    // In 1000 ms remove the 'active' class
+	    asyncDelay(1000).then(() => copyIcon.classList.remove('active'))
+            const uiState = readUI()
+            TASK_REQ_NO_EXPORT.forEach((key) => delete uiState.reqBody[key])
+            if (uiState.reqBody.init_image && !IMAGE_REGEX.test(uiState.reqBody.init_image)) {
+                delete uiState.reqBody.init_image
+                delete uiState.reqBody.prompt_strength
+            }
+            navigator.clipboard.writeText(JSON.stringify(uiState, undefined, 4))
+        })
+        resetSettings.parentNode.insertBefore(copyIcon, resetSettings)
+
+        // PASTE ICON
+        const pasteIcon = document.createElement('i')
+        pasteIcon.className = 'fa-solid fa-paste section-button'
+        pasteIcon.innerHTML = `<span class="simple-tooltip right">Paste Image Settings</span>`
+        pasteIcon.addEventListener('click', (event) => {
+            event.stopPropagation()
+	    // Add css class 'active'
+	    pasteIcon.classList.add('active')
+	    // In 1000 ms remove the 'active' class
+	    asyncDelay(1000).then(() => pasteIcon.classList.remove('active'))
+            pasteFromClipboard()
+        })
+        resetSettings.parentNode.insertBefore(pasteIcon, resetSettings)
+    }
+}
+
+// Determine which access we have to the clipboard. Clipboard access is only available  on localhost or via TLS.
+navigator.permissions.query({ name: "clipboard-write" }).then(checkWriteToClipboardPermission, (e) => {
+    if (e instanceof TypeError && typeof navigator?.clipboard?.writeText === 'function') {
+        // Fix for firefox https://bugzilla.mozilla.org/show_bug.cgi?id=1560373
+        checkWriteToClipboardPermission({state:"granted"})
+    }
+})
--- a/ui/media/js/image-modifiers.js
+++ b/ui/media/js/image-modifiers.js
@ -75,7 +75,6 @@ function createModifierGroup(modifierGroup, initiallyExpanded) {

    if (initiallyExpanded === true) {
        titleEl.className += ' active'
-        modifiersEl.style.display = 'block'
    }

    modifiers.forEach(modObj => {
@ -245,16 +244,9 @@ function resizeModifierCards(val) {
 modifierCardSizeSlider.onchange = () => resizeModifierCards(modifierCardSizeSlider.value)
 previewImageField.onchange = () => changePreviewImages(previewImageField.value)

-modifierSettingsBtn.addEventListener('click', function() {
-    modifierSettingsOverlay.style.display = 'block'
-})
-document.getElementById("modifier-settings-config-close-btn").addEventListener('click', () => {
-    modifierSettingsOverlay.style.display = 'none'
-})
-modifierSettingsOverlay.addEventListener('click', (event) => {
-    if (event.target.id == modifierSettingsOverlay.id) {
-        modifierSettingsOverlay.style.display = 'none'
-    }
+modifierSettingsBtn.addEventListener('click', function(e) {
+    modifierSettingsOverlay.classList.add("active")
+    e.stopPropagation()
 })

 function saveCustomModifiers() {
--- a/ui/media/js/main.js
+++ b/ui/media/js/main.js
@ -1,6 +1,7 @@
 "use strict" // Opt in to a restricted variant of JavaScript
 const HEALTH_PING_INTERVAL = 5 // seconds
 const MAX_INIT_IMAGE_DIMENSION = 768
+const MIN_GPUS_TO_SHOW_SELECTION = 2

 const IMAGE_REGEX = new RegExp('data:image/[A-Za-z]+;base64')

@ -24,13 +25,6 @@ let initImagePreview = document.querySelector("#init_image_preview")
 let initImageSizeBox = document.querySelector("#init_image_size_box")
 let maskImageSelector = document.querySelector("#mask")
 let maskImagePreview = document.querySelector("#mask_preview")
-let turboField = document.querySelector('#turbo')
-let useCPUField = document.querySelector('#use_cpu')
-let useFullPrecisionField = document.querySelector('#use_full_precision')
-let saveToDiskField = document.querySelector('#save_to_disk')
-let diskPathField = document.querySelector('#diskPath')
-// let allowNSFWField = document.querySelector("#allow_nsfw")
-let useBetaChannelField = document.querySelector("#use_beta_channel")
 let promptStrengthSlider = document.querySelector('#prompt_strength_slider')
 let promptStrengthField = document.querySelector('#prompt_strength')
 let samplerField = document.querySelector('#sampler')
@ -39,6 +33,7 @@ let useFaceCorrectionField = document.querySelector("#use_face_correction")
 let useUpscalingField = document.querySelector("#use_upscale")
 let upscaleModelField = document.querySelector("#upscale_model")
 let stableDiffusionModelField = document.querySelector('#stable_diffusion_model')
+let vaeModelField = document.querySelector('#vae_model')
 let outputFormatField = document.querySelector('#output_format')
 let showOnlyFilteredImageField = document.querySelector("#show_only_filtered_image")
 let updateBranchLabel = document.querySelector("#updateBranchLabel")
@ -56,22 +51,10 @@ let initialText = document.querySelector("#initial-text")
 let previewTools = document.querySelector("#preview-tools")
 let clearAllPreviewsBtn = document.querySelector("#clear-all-previews")

-// let maskSetting = document.querySelector('#editor-inputs-mask_setting')
-// let maskImagePreviewContainer = document.querySelector('#mask_preview_container')
-// let maskImageClearBtn = document.querySelector('#mask_clear')
 let maskSetting = document.querySelector('#enable_mask')

 let imagePreview = document.querySelector("#preview")

-// let previewPrompt = document.querySelector('#preview-prompt')
-
-let showConfigToggle = document.querySelector('#configToggleBtn')
-// let configBox = document.querySelector('#config')
-// let outputMsg = document.querySelector('#outputMsg')
-// let progressBar = document.querySelector("#progressBar")
-
-let soundToggle = document.querySelector('#sound_toggle')
-
 let serverStatusColor = document.querySelector('#server-status-color')
 let serverStatusMsg = document.querySelector('#server-status-msg')

@ -85,7 +68,6 @@ maskResetButton.style.fontWeight = 'normal'
 maskResetButton.style.fontSize = '10pt'

 let serverState = {'status': 'Offline', 'time': Date.now()}
-let lastPromptUsed = ''
 let bellPending = false

 let taskQueue = []
@ -187,6 +169,34 @@ function playSound() {
        })
    }
 }
+function setSystemInfo(devices) {
+    let cpu = devices.all.cpu.name
+    let allGPUs = Object.keys(devices.all).filter(d => d != 'cpu')
+    let activeGPUs = Object.keys(devices.active)
+
+    function ID_TO_TEXT(d) {
+        let info = devices.all[d]
+        if ("mem_free" in info && "mem_total" in info) {
+            return `${info.name} <small>(${d}) (${info.mem_free.toFixed(1)}Gb free / ${info.mem_total.toFixed(1)} Gb total)</small>`
+        } else {
+            return `${info.name} <small>(${d}) (no memory info)</small>`
+        }
+    }
+
+    allGPUs = allGPUs.map(ID_TO_TEXT)
+    activeGPUs = activeGPUs.map(ID_TO_TEXT)
+
+    let systemInfo = `
+    <table>
+        <tr><td><label>Processor:</label></td><td class="value">${cpu}</td></tr>
+        <tr><td><label>Compatible Graphics Cards (all):</label></td><td class="value">${allGPUs.join('</br>')}</td></tr>
+        <tr><td></td><td>&nbsp;</td></tr>
+        <tr><td><label>Used for rendering 🔥:</label></td><td class="value">${activeGPUs.join('</br>')}</td></tr>
+    </table>`
+
+    let systemInfoEl = document.querySelector('#system-info')
+    systemInfoEl.innerHTML = systemInfo
+}

 async function healthCheck() {
    try {
@ -220,8 +230,12 @@ async function healthCheck() {
                setServerStatus('error', serverState.status.toLowerCase())
                break
        }
+        if (serverState.devices) {
+            setSystemInfo(serverState.devices)
+        }
        serverState.time = Date.now()
    } catch (e) {
+        console.log(e)
        serverState = {'status': 'Offline', 'time': Date.now()}
        setServerStatus('error', 'offline')
    }
@ -340,10 +354,10 @@ function onDownloadImageClick(req, img) {
    imgDownload.click()
 }

-function modifyCurrentRequest(req, ...reqDiff) {
+function modifyCurrentRequest(...reqDiff) {
    const newTaskRequest = getCurrentUserRequest()

-    newTaskRequest.reqBody = Object.assign({}, req, ...reqDiff, {
+    newTaskRequest.reqBody = Object.assign(newTaskRequest.reqBody, ...reqDiff, {
        use_cpu: useCPUField.checked
    })
    newTaskRequest.seed = newTaskRequest.reqBody.seed
@ -410,7 +424,7 @@ async function doMakeImage(task) {
    const RETRY_DELAY_IF_BUFFER_IS_EMPTY = 1000 // ms
    const RETRY_DELAY_IF_SERVER_IS_BUSY = 30 * 1000 // ms, status_code 503, already a task running
    const TASK_START_DELAY_ON_SERVER = 1500 // ms
-    const SERVER_STATE_VALIDITY_DURATION = 10 * 1000 // ms
+    const SERVER_STATE_VALIDITY_DURATION = 90 * 1000 // ms

    const reqBody = task.reqBody
    const batchCount = task.batchCount
@ -422,10 +436,10 @@ async function doMakeImage(task) {
    const outputMsg = task['outputMsg']
    const previewPrompt = task['previewPrompt']
    const progressBar = task['progressBar']
+    const progressBarInner = progressBar.querySelector("div")

    let res = undefined
    try {
-        const lastTask = serverState.task
        let renderRequest = undefined
        do {
            res = await fetch('/render', {
@ -561,6 +575,13 @@ async function doMakeImage(task) {
                outputMsg.innerHTML += `. Time remaining (approx): ${timeRemaining}`
                outputMsg.style.display = 'block'

+                progressBarInner.style.width = `${percent}%`
+                if (percent == 100) {
+                    task.progressBar.style.height = "0px"
+                    task.progressBar.style.border = "0px solid var(--background-color3)"
+                    task.progressBar.classList.remove("active")
+                }
+
                if (stepUpdate.output !== undefined) {
                    showImages(reqBody, stepUpdate, outputContainer, true)
                }
@ -620,17 +641,14 @@ async function doMakeImage(task) {
                let msg = `Unexpected Read Error:<br/><pre>Response: ${res}<br/>StepUpdate: ${typeof stepUpdate === 'object' ? JSON.stringify(stepUpdate, undefined, 4) : stepUpdate}</pre>`
                logError(msg, res, outputMsg)
            }
-            progressBar.style.display = 'none'
            return false
        }

-        lastPromptUsed = reqBody['prompt']
        showImages(reqBody, stepUpdate, outputContainer, false)
    } catch (e) {
        console.log('request error', e)
        logError('Stable Diffusion had an error. Please check the logs in the command-line window. <br/><br/>' + e + '<br/><pre>' + e.stack + '</pre>', res, outputMsg)
        setStatus('request', 'error', 'error')
-        progressBar.style.display = 'none'
        return false
    }
    return true
@ -713,6 +731,9 @@ async function checkTasks() {

    if (successCount === task.batchCount) {
        task.outputMsg.innerText = 'Processed ' + task.numOutputsTotal + ' images in ' + time + ' seconds'
+        task.progressBar.style.height = "0px"
+        task.progressBar.style.border = "0px solid var(--background-color3)"
+        task.progressBar.classList.remove("active")
        // setStatus('request', 'done', 'success')
    } else {
        if (task.outputMsg.innerText.toLowerCase().indexOf('error') === -1) {
@ -762,9 +783,9 @@ function getCurrentUserRequest() {
            height: heightField.value,
            // allow_nsfw: allowNSFWField.checked,
            turbo: turboField.checked,
-            use_cpu: useCPUField.checked,
            use_full_precision: useFullPrecisionField.checked,
            use_stable_diffusion_model: stableDiffusionModelField.value,
+            use_vae_model: vaeModelField.value,
            stream_progress_updates: true,
            stream_image_progress: (numOutputsTotal > 50 ? false : streamImageProgressField.checked),
            show_only_filtered_image: showOnlyFilteredImageField.checked,
@ -813,29 +834,34 @@ function makeImage() {
 }

 function createTask(task) {
-    let taskConfig = `Seed: ${task.seed}, Sampler: ${task.reqBody.sampler}, Inference Steps: ${task.reqBody.num_inference_steps}, Guidance Scale: ${task.reqBody.guidance_scale}, Model: ${task.reqBody.use_stable_diffusion_model}`
-    if (negativePromptField.value.trim() !== '') {
-        taskConfig += `, Negative Prompt: ${task.reqBody.negative_prompt}`
+    let taskConfig = `<b>Seed:</b> ${task.seed}, <b>Sampler:</b> ${task.reqBody.sampler}, <b>Inference Steps:</b> ${task.reqBody.num_inference_steps}, <b>Guidance Scale:</b> ${task.reqBody.guidance_scale}, <b>Model:</b> ${task.reqBody.use_stable_diffusion_model}`
+    if (task.reqBody.use_vae_model.trim() !== '') {
+        taskConfig += `, <b>VAE:</b> ${task.reqBody.use_vae_model}`
+    }
+    if (task.reqBody.negative_prompt.trim() !== '') {
+        taskConfig += `, <b>Negative Prompt:</b> ${task.reqBody.negative_prompt}`
    }
    if (task.reqBody.init_image !== undefined) {
-        taskConfig += `, Prompt Strength: ${task.reqBody.prompt_strength}`
+        taskConfig += `, <b>Prompt Strength:</b> ${task.reqBody.prompt_strength}`
    }
    if (task.reqBody.use_face_correction) {
-        taskConfig += `, Fix Faces: ${task.reqBody.use_face_correction}`
+        taskConfig += `, <b>Fix Faces:</b> ${task.reqBody.use_face_correction}`
    }
    if (task.reqBody.use_upscale) {
-        taskConfig += `, Upscale: ${task.reqBody.use_upscale}`
+        taskConfig += `, <b>Upscale:</b> ${task.reqBody.use_upscale}`
    }

    let taskEntry = document.createElement('div')
    taskEntry.className = 'imageTaskContainer'
-    taskEntry.innerHTML = ` <div class="taskStatusLabel">Enqueued</div>
-                            <button class="secondaryButton stopTask"><i class="fa-solid fa-trash-can"></i> Remove</button>
-                            <div class="preview-prompt collapsible active"></div>
-                            <div class="taskConfig">${taskConfig}</div>
-                            <div class="collapsible-content" style="display: block">
+    taskEntry.innerHTML = ` <div class="header-content panel collapsible active">
+                                <div class="taskStatusLabel">Enqueued</div>
+                                <button class="secondaryButton stopTask"><i class="fa-solid fa-trash-can"></i> Remove</button>
+                                <div class="preview-prompt collapsible active"></div>
+                                <div class="taskConfig">${taskConfig}</div>
                                <div class="outputMsg"></div>
-                                <div class="progressBar"></div>
+                                <div class="progress-bar active"><div></div></div>
+                            </div>
+                            <div class="collapsible-content">
                                <div class="img-preview">
                            </div>`

@ -845,12 +871,14 @@ function createTask(task) {
    task['outputContainer'] = taskEntry.querySelector('.img-preview')
    task['outputMsg'] = taskEntry.querySelector('.outputMsg')
    task['previewPrompt'] = taskEntry.querySelector('.preview-prompt')
-    task['progressBar'] = taskEntry.querySelector('.progressBar')
+    task['progressBar'] = taskEntry.querySelector('.progress-bar')
    task['stopTask'] = taskEntry.querySelector('.stopTask')

-    task['stopTask'].addEventListener('click', async function() {
+    task['stopTask'].addEventListener('click', async function(e) {
+        e.stopPropagation()
        if (task['isProcessing']) {
            task.isProcessing = false
+            task.progressBar.classList.remove("active")
            try {
                let res = await fetch('/image/stop?session_id=' + sessionId)
            } catch (e) {
@ -1044,16 +1072,25 @@ function onDimensionChange() {
    resizeInpaintingEditor(widthValue, heightValue)
 }

-saveToDiskField.addEventListener('click', function(e) {
-    diskPathField.disabled = !this.checked
-})
+diskPathField.disabled = !saveToDiskField.checked

-useUpscalingField.addEventListener('click', function(e) {
+upscaleModelField.disabled = !useUpscalingField.checked
+useUpscalingField.addEventListener('change', function(e) {
    upscaleModelField.disabled = !this.checked
 })

+if (useBetaChannelField.checked) {
+    updateBranchLabel.innerText = "(beta)"
+}
+
 makeImageBtn.addEventListener('click', makeImage)

+document.onkeydown = function(e) {
+    if (e.ctrlKey && e.code === 'Enter') {
+        makeImage()
+        e.preventDefault()
+    }
+}

 function updateGuidanceScale() {
    guidanceScaleField.value = guidanceScaleSlider.value / 10
@ -1095,80 +1132,44 @@ promptStrengthSlider.addEventListener('input', updatePromptStrength)
 promptStrengthField.addEventListener('input', updatePromptStrengthSlider)
 updatePromptStrength()

-useBetaChannelField.addEventListener('click', async function(e) {
-    if (!isServerAvailable()) {
-        // logError('The server is still starting up..')
-        alert('The server is still starting up..')
-        e.preventDefault()
-        return false
-    }
-
-    let updateBranch = (this.checked ? 'beta' : 'main')
-
-    try {
-        let res = await fetch('/app_config', {
-            method: 'POST',
-            headers: {
-                'Content-Type': 'application/json'
-            },
-            body: JSON.stringify({
-                'update_branch': updateBranch
-            })
-        })
-        res = await res.json()
-
-        console.log('set config status response', res)
-    } catch (e) {
-        console.log('set config status error', e)
-    }
-})
-
-async function getAppConfig() {
-    try {
-        let res = await fetch('/get/app_config')
-        const config = await res.json()
-
-        if (config.update_branch === 'beta') {
-            useBetaChannelField.checked = true
-            updateBranchLabel.innerText = "(beta)"
-        }
-
-        console.log('get config status response', config)
-    } catch (e) {
-        console.log('get config status error', e)
-    }
-}
-
 async function getModels() {
    try {
-        var model_setting_key = "stable_diffusion_model"
-        var selectedModel = SETTINGS[model_setting_key].value
+        var sd_model_setting_key = "stable_diffusion_model"
+        var vae_model_setting_key = "vae_model"
+        var selectedSDModel = SETTINGS[sd_model_setting_key].value
+        var selectedVaeModel = SETTINGS[vae_model_setting_key].value
        let res = await fetch('/get/models')
        const models = await res.json()

-        // let activeModel = models['active']
+        console.log('get models response', models)
+
        let modelOptions = models['options']
        let stableDiffusionOptions = modelOptions['stable-diffusion']
+        let vaeOptions = modelOptions['vae']
+        vaeOptions.unshift('') // add a None option

-        stableDiffusionOptions.forEach(modelName => {
-            let modelOption = document.createElement('option')
-            modelOption.value = modelName
-            modelOption.innerText = modelName
+        function createModelOptions(modelField, selectedModel) {
+            return function(modelName) {
+                let modelOption = document.createElement('option')
+                modelOption.value = modelName
+                modelOption.innerText = modelName !== '' ? modelName : 'None'

-            if (modelName === selectedModel) {
-                modelOption.selected = true
+                if (modelName === selectedModel) {
+                    modelOption.selected = true
+                }
+
+                modelField.appendChild(modelOption)
            }
-
-            stableDiffusionModelField.appendChild(modelOption)
-        })
-
-        // TODO: set default for model here too
-        SETTINGS[model_setting_key].default = stableDiffusionOptions[0]
-        if (getSetting(model_setting_key) == '' || SETTINGS[model_setting_key].value == '') {
-            setSetting(model_setting_key, stableDiffusionOptions[0])
        }

-        console.log('get models response', models)
+        stableDiffusionOptions.forEach(createModelOptions(stableDiffusionModelField, selectedSDModel))
+        vaeOptions.forEach(createModelOptions(vaeModelField, selectedVaeModel))
+
+        // TODO: set default for model here too
+        SETTINGS[sd_model_setting_key].default = stableDiffusionOptions[0]
+        if (getSetting(sd_model_setting_key) == '' || SETTINGS[sd_model_setting_key].value == '') {
+            setSetting(sd_model_setting_key, stableDiffusionOptions[0])
+        }
    } catch (e) {
        console.log('get models error', e)
    }
@ -1267,21 +1268,56 @@ promptsFromFileSelector.addEventListener('change', function() {
    }
 })

-async function getDiskPath() {
-    try {
-        var diskPath = getSetting("diskPath")
-        if (diskPath == '' || diskPath == undefined || diskPath == "undefined") {
-            let res = await fetch('/get/output_dir')
-            if (res.status === 200) {
-                res = await res.json()
-                res = res.output_dir
-
-                setSetting("diskPath", res)
-            }
+/* setup popup handlers */
+document.querySelectorAll('.popup').forEach(popup => {
+    popup.addEventListener('click', event => {
+        if (event.target == popup) {
+            popup.classList.remove("active")
        }
-    } catch (e) {
-        console.log('error fetching output dir path', e)
+    })
+    var closeButton = popup.querySelector(".close-button")
+    if (closeButton) {
+        closeButton.addEventListener('click', () => {
+            popup.classList.remove("active")
+        })
    }
-}
+})
+
+var tabElements = [];
+document.querySelectorAll(".tab").forEach(tab => {
+    var name = tab.id.replace("tab-", "");
+    var content = document.getElementById(`tab-content-${name}`)
+    tabElements.push({
+        name: name,
+        tab: tab,
+        content: content
+    })
+
+    tab.addEventListener("click", event => {
+        if (!tab.classList.contains("active")) {
+            tabElements.forEach(tabInfo => {
+                if (tabInfo.tab.classList.contains("active")) {
+                    tabInfo.tab.classList.toggle("active")
+                    tabInfo.content.classList.toggle("active")
+                }
+            })
+            tab.classList.toggle("active")
+            content.classList.toggle("active")
+        }
+    })
+})
+
+window.addEventListener("beforeunload", function(e) {
+    const msg = "Unsaved pictures will be lost!";
+
+    let elementList = document.getElementsByClassName("imageTaskContainer");
+    if (elementList.length != 0) {
+        e.preventDefault();
+        (e || window.event).returnValue = msg;
+        return msg;
+    } else {
+        return true;
+    }
+});

 createCollapsibles()
--- a/ui/media/js/parameters.js
+++ b/ui/media/js/parameters.js
@ -0,0 +1,330 @@
+/**
+ * Enum of parameter types
+ * @readonly
+ * @enum {string}
+ */
+ var ParameterType = {
+    checkbox: "checkbox",
+	select: "select",
+	select_multiple: "select_multiple",
+	custom: "custom",
+};
+
+/**
+ * JSDoc style
+ * @typedef {object} Parameter
+ * @property {string} id
+ * @property {ParameterType} type
+ * @property {string} label
+ * @property {?string} note
+ * @property {number|boolean|string} default
+ */
+
+
+/** @type {Array.<Parameter>} */
+var PARAMETERS = [
+	{
+		id: "theme",
+		type: ParameterType.select,
+		label: "Theme",
+		default: "theme-default",
+		options: [ // Note: options expanded dynamically
+			{
+				value: "theme-default",
+				label: "Default"
+			}
+		]
+	},
+	{
+		id: "save_to_disk",
+		type: ParameterType.checkbox,
+		label: "Auto-Save Images",
+		note: "automatically saves images to the specified location",
+		default: false,
+	},
+	{
+		id: "diskPath",
+		type: ParameterType.custom,
+		label: "Save Location",
+		render: (parameter) => {
+			return `<input id="${parameter.id}" name="${parameter.id}" size="30" disabled>`
+		}
+	},
+	{
+		id: "sound_toggle",
+		type: ParameterType.checkbox,
+		label: "Enable Sound",
+		note: "plays a sound on task completion",
+		default: true,
+	},
+	{
+		id: "ui_open_browser_on_start",
+		type: ParameterType.checkbox,
+		label: "Open browser on startup",
+		note: "starts the default browser on startup",
+		default: true,
+	},
+	{
+		id: "turbo",
+		type: ParameterType.checkbox,
+		label: "Turbo Mode",
+		default: true,
+		note: "generates images faster, but uses an additional 1 GB of GPU memory",
+	},
+	{
+		id: "use_cpu",
+		type: ParameterType.checkbox,
+		label: "Use CPU (not GPU)",
+		note: "warning: this will be *very* slow",
+		default: false,
+	},
+	{
+		id: "auto_pick_gpus",
+		type: ParameterType.checkbox,
+		label: "Automatically pick the GPUs (experimental)",
+		default: false,
+	},
+	{
+		id: "use_gpus",
+		type: ParameterType.select_multiple,
+		label: "GPUs to use (experimental)",
+		note: "to process in parallel",
+		default: false,
+	},
+	{
+		id: "use_full_precision",
+		type: ParameterType.checkbox,
+		label: "Use Full Precision",
+		note: "for GPU-only. warning: this will consume more VRAM",
+		default: false,
+	},
+	{
+		id: "auto_save_settings",
+		type: ParameterType.checkbox,
+		label: "Auto-Save Settings",
+		note: "restores settings on browser load",
+		default: true,
+	},
+	{
+		id: "use_beta_channel",
+		type: ParameterType.checkbox,
+		label: "🔥Beta channel",
+		note: "Get the latest features immediately (but could be less stable). Please restart the program after changing this.",
+		default: false,
+	},
+];
+
+function getParameterSettingsEntry(id) {
+	let parameter = PARAMETERS.filter(p => p.id === id)
+	if (parameter.length === 0) {
+		return
+	}
+	return parameter[0].settingsEntry
+}
+
+function getParameterElement(parameter) {
+	switch (parameter.type) {
+		case ParameterType.checkbox:
+			var is_checked = parameter.default ? " checked" : "";
+			return `<input id="${parameter.id}" name="${parameter.id}"${is_checked} type="checkbox">`
+		case ParameterType.select:
+		case ParameterType.select_multiple:
+			var options = (parameter.options || []).map(option => `<option value="${option.value}">${option.label}</option>`).join("")
+			var multiple = (parameter.type == ParameterType.select_multiple ? 'multiple' : '')
+			return `<select id="${parameter.id}" name="${parameter.id}" ${multiple}>${options}</select>`
+		case ParameterType.custom:
+			return parameter.render(parameter)
+		default:
+			console.error(`Invalid type for parameter ${parameter.id}`);
+			return "ERROR: Invalid Type"
+	}
+}
+
+let parametersTable = document.querySelector("#system-settings table")
+/* fill in the system settings popup table */
+function initParameters() {
+	PARAMETERS.forEach(parameter => {
+		var element = getParameterElement(parameter)
+		var note = parameter.note ? `<small>${parameter.note}</small>` : "";
+		var newrow = document.createElement('tr')
+		newrow.innerHTML = `
+			<td><label for="${parameter.id}">${parameter.label}</label></td>
+			<td><div>${element}${note}<div></td>`
+		parametersTable.appendChild(newrow)
+		parameter.settingsEntry = newrow
+	})
+}
+
+initParameters()
+
+let turboField = document.querySelector('#turbo')
+let useCPUField = document.querySelector('#use_cpu')
+let autoPickGPUsField = document.querySelector('#auto_pick_gpus')
+let useGPUsField = document.querySelector('#use_gpus')
+let useFullPrecisionField = document.querySelector('#use_full_precision')
+let saveToDiskField = document.querySelector('#save_to_disk')
+let diskPathField = document.querySelector('#diskPath')
+let useBetaChannelField = document.querySelector("#use_beta_channel")
+let uiOpenBrowserOnStartField = document.querySelector("#ui_open_browser_on_start")
+
+let saveSettingsBtn = document.querySelector('#save-system-settings-btn')
+
+async function changeAppConfig(configDelta) {
+    try {
+        let res = await fetch('/app_config', {
+            method: 'POST',
+            headers: {
+                'Content-Type': 'application/json'
+            },
+            body: JSON.stringify(configDelta)
+        })
+        res = await res.json()
+
+        console.log('set config status response', res)
+    } catch (e) {
+        console.log('set config status error', e)
+    }
+}
+
+async function getAppConfig() {
+    try {
+        let res = await fetch('/get/app_config')
+        const config = await res.json()
+
+        if (config.update_branch === 'beta') {
+            useBetaChannelField.checked = true
+        }
+        if (config.ui && config.ui.open_browser_on_start === false) {
+            uiOpenBrowserOnStartField.checked = false
+        }
+
+        console.log('get config status response', config)
+    } catch (e) {
+        console.log('get config status error', e)
+    }
+}
+
+saveToDiskField.addEventListener('change', function(e) {
+    diskPathField.disabled = !this.checked
+})
+
+function getCurrentRenderDeviceSelection() {
+    let selectedGPUs = $('#use_gpus').val()
+
+    if (useCPUField.checked && !autoPickGPUsField.checked) {
+        return 'cpu'
+    }
+    if (autoPickGPUsField.checked || selectedGPUs.length == 0) {
+        return 'auto'
+    }
+
+    return selectedGPUs.join(',')
+}
+
+useCPUField.addEventListener('click', function() {
+    let gpuSettingEntry = getParameterSettingsEntry('use_gpus')
+    let autoPickGPUSettingEntry = getParameterSettingsEntry('auto_pick_gpus')
+    if (this.checked) {
+        gpuSettingEntry.style.display = 'none'
+        autoPickGPUSettingEntry.style.display = 'none'
+        autoPickGPUsField.setAttribute('data-old-value', autoPickGPUsField.checked)
+        autoPickGPUsField.checked = false
+    } else if (useGPUsField.options.length >= MIN_GPUS_TO_SHOW_SELECTION) {
+        gpuSettingEntry.style.display = ''
+        autoPickGPUSettingEntry.style.display = ''
+        let oldVal = autoPickGPUsField.getAttribute('data-old-value')
+        if (oldVal === null || oldVal === undefined) { // the UI started with CPU selected by default
+            autoPickGPUsField.checked = true
+        } else {
+            autoPickGPUsField.checked = (oldVal === 'true')
+        }
+        gpuSettingEntry.style.display = (autoPickGPUsField.checked ? 'none' : '')
+    }
+})
+
+useGPUsField.addEventListener('click', function() {
+    let selectedGPUs = $('#use_gpus').val()
+    autoPickGPUsField.checked = (selectedGPUs.length === 0)
+})
+
+autoPickGPUsField.addEventListener('click', function() {
+    if (this.checked) {
+        $('#use_gpus').val([])
+    }
+
+    let gpuSettingEntry = getParameterSettingsEntry('use_gpus')
+    gpuSettingEntry.style.display = (this.checked ? 'none' : '')
+})
+
+async function getDiskPath() {
+    try {
+        var diskPath = getSetting("diskPath")
+        if (diskPath == '' || diskPath == undefined || diskPath == "undefined") {
+            let res = await fetch('/get/output_dir')
+            if (res.status === 200) {
+                res = await res.json()
+                res = res.output_dir
+
+                setSetting("diskPath", res)
+            }
+        }
+    } catch (e) {
+        console.log('error fetching output dir path', e)
+    }
+}
+
+async function getDevices() {
+    try {
+        let res = await fetch('/get/devices')
+        if (res.status === 200) {
+            res = await res.json()
+
+            let allDeviceIds = Object.keys(res['all']).filter(d => d !== 'cpu')
+            let activeDeviceIds = Object.keys(res['active']).filter(d => d !== 'cpu')
+
+            if (activeDeviceIds.length === 0) {
+                useCPUField.checked = true
+            }
+
+            if (allDeviceIds.length < MIN_GPUS_TO_SHOW_SELECTION || useCPUField.checked) {
+                let gpuSettingEntry = getParameterSettingsEntry('use_gpus')
+                gpuSettingEntry.style.display = 'none'
+                let autoPickGPUSettingEntry = getParameterSettingsEntry('auto_pick_gpus')
+                autoPickGPUSettingEntry.style.display = 'none'
+            }
+
+            if (allDeviceIds.length === 0) {
+                useCPUField.checked = true
+                useCPUField.disabled = true // no compatible GPUs, so make the CPU mandatory
+            }
+
+            autoPickGPUsField.checked = (res['config'] === 'auto')
+
+            useGPUsField.innerHTML = ''
+            allDeviceIds.forEach(device => {
+                let deviceName = res['all'][device]['name']
+                let deviceOption = `<option value="${device}">${deviceName} (${device})</option>`
+                useGPUsField.insertAdjacentHTML('beforeend', deviceOption)
+            })
+
+            if (autoPickGPUsField.checked) {
+                let gpuSettingEntry = getParameterSettingsEntry('use_gpus')
+                gpuSettingEntry.style.display = 'none'
+            } else {
+                $('#use_gpus').val(activeDeviceIds)
+            }
+        }
+    } catch (e) {
+        console.log('error fetching devices', e)
+    }
+}
+
+saveSettingsBtn.addEventListener('click', function() {
+	let updateBranch = (useBetaChannelField.checked ? 'beta' : 'main')
+
+	changeAppConfig({
+        'render_devices': getCurrentRenderDeviceSelection(),
+		'update_branch': updateBranch,
+		'ui_open_browser_on_start': uiOpenBrowserOnStartField.checked
+    })
+})
--- a/ui/media/js/utils.js
+++ b/ui/media/js/utils.js
@ -29,12 +29,16 @@ function toggleCollapsible(element) {
    var handle = element.querySelector(".collapsible-handle");
    collapsibleHeader.classList.toggle("active")
    let content = getNextSibling(collapsibleHeader, '.collapsible-content')
-    if (content.style.display === "block") {
+    if (!collapsibleHeader.classList.contains("active")) {
        content.style.display = "none"
-        handle.innerHTML = '&#x2795;' // plus
+        if (handle != null) {  // render results don't have a handle
+            handle.innerHTML = '&#x2795;' // plus
+        }
    } else {
        content.style.display = "block"
-        handle.innerHTML = '&#x2796;' // minus
+        if (handle != null) {  // render results don't have a handle
+            handle.innerHTML = '&#x2796;' // minus
+        }
    }
    
    if (COLLAPSIBLES_INITIALIZED && COLLAPSIBLE_PANELS.includes(element)) {
@ -65,7 +69,7 @@ function createCollapsibles(node) {
        let handle = document.createElement('span')
        handle.className = 'collapsible-handle'

-        if (c.className.indexOf('active') !== -1) {
+        if (c.classList.contains("active")) {
            handle.innerHTML = '&#x2796;' // minus
        } else {
            handle.innerHTML = '&#x2795;' // plus
--- a/ui/sd_internal/init.py
+++ b/ui/sd_internal/init.py
@ -18,11 +18,11 @@ class Request:
    precision: str = "autocast" # or "full"
    save_to_disk_path: str = None
    turbo: bool = True
-    use_cpu: bool = False
    use_full_precision: bool = False
    use_face_correction: str = None # or "GFPGANv1.3"
    use_upscale: str = None # or "RealESRGAN_x4plus" or "RealESRGAN_x4plus_anime_6B"
    use_stable_diffusion_model: str = "sd-v1-4"
+    use_vae_model: str = None
    show_only_filtered_image: bool = False
    output_format: str = "jpeg" # or "png"

@ -45,10 +45,11 @@ class Request:
            "use_face_correction": self.use_face_correction,
            "use_upscale": self.use_upscale,
            "use_stable_diffusion_model": self.use_stable_diffusion_model,
+            "use_vae_model": self.use_vae_model,
            "output_format": self.output_format,
        }

-    def to_string(self):
+    def __str__(self):
        return f'''
    session_id: {self.session_id}
    prompt: {self.prompt}
@ -62,11 +63,11 @@ class Request:
    precision: {self.precision}
    save_to_disk_path: {self.save_to_disk_path}
    turbo: {self.turbo}
-    use_cpu: {self.use_cpu}
    use_full_precision: {self.use_full_precision}
    use_face_correction: {self.use_face_correction}
    use_upscale: {self.use_upscale}
    use_stable_diffusion_model: {self.use_stable_diffusion_model}
+    use_vae_model: {self.use_vae_model}
    show_only_filtered_image: {self.show_only_filtered_image}
    output_format: {self.output_format}

--- a/ui/sd_internal/device_manager.py
+++ b/ui/sd_internal/device_manager.py
@ -0,0 +1,168 @@
+import os
+import torch
+import traceback
+import re
+
+COMPARABLE_GPU_PERCENTILE = 0.65 # if a GPU's free_mem is within this % of the GPU with the most free_mem, it will be picked
+
+mem_free_threshold = 0
+
+def get_device_delta(render_devices, active_devices):
+    '''
+    render_devices: 'cpu', or 'auto' or ['cuda:N'...]
+    active_devices: ['cpu', 'cuda:N'...]
+    '''
+
+    if render_devices in ('cpu', 'auto'):
+        render_devices = [render_devices]
+    elif render_devices is not None:
+        if isinstance(render_devices, str):
+            render_devices = [render_devices]
+        if isinstance(render_devices, list) and len(render_devices) > 0:
+            render_devices = list(filter(lambda x: x.startswith('cuda:'), render_devices))
+            if len(render_devices) == 0:
+                raise Exception('Invalid render_devices value in config.json. Valid: {"render_devices": ["cuda:0", "cuda:1"...]}, or {"render_devices": "cpu"} or {"render_devices": "auto"}')
+
+            render_devices = list(filter(lambda x: is_device_compatible(x), render_devices))
+            if len(render_devices) == 0:
+                raise Exception('Sorry, none of the render_devices configured in config.json are compatible with Stable Diffusion')
+        else:
+            raise Exception('Invalid render_devices value in config.json. Valid: {"render_devices": ["cuda:0", "cuda:1"...]}, or {"render_devices": "cpu"} or {"render_devices": "auto"}')
+    else:
+        render_devices = ['auto']
+
+    if 'auto' in render_devices:
+        render_devices = auto_pick_devices(active_devices)
+        if 'cpu' in render_devices:
+            print('WARNING: Could not find a compatible GPU. Using the CPU, but this will be very slow!')
+
+    active_devices = set(active_devices)
+    render_devices = set(render_devices)
+
+    devices_to_start = render_devices - active_devices
+    devices_to_stop = active_devices - render_devices
+
+    return devices_to_start, devices_to_stop
+
+def auto_pick_devices(currently_active_devices):
+    global mem_free_threshold
+
+    if not torch.cuda.is_available(): return ['cpu']
+
+    device_count = torch.cuda.device_count()
+    if device_count == 1:
+        return ['cuda:0'] if is_device_compatible('cuda:0') else ['cpu']
+
+    print('Autoselecting GPU. Using most free memory.')
+    devices = []
+    for device in range(device_count):
+        device = f'cuda:{device}'
+        if not is_device_compatible(device):
+            continue
+
+        mem_free, mem_total = torch.cuda.mem_get_info(device)
+        mem_free /= float(10**9)
+        mem_total /= float(10**9)
+        device_name = torch.cuda.get_device_name(device)
+        print(f'{device} detected: {device_name} - Memory (free/total): {round(mem_free, 2)}Gb / {round(mem_total, 2)}Gb')
+        devices.append({'device': device, 'device_name': device_name, 'mem_free': mem_free})
+
+    devices.sort(key=lambda x:x['mem_free'], reverse=True)
+    max_mem_free = devices[0]['mem_free']
+    curr_mem_free_threshold = COMPARABLE_GPU_PERCENTILE * max_mem_free
+    mem_free_threshold = max(curr_mem_free_threshold, mem_free_threshold)
+
+    # Auto-pick algorithm:
+    # 1. Pick the top 75 percentile of the GPUs, sorted by free_mem.
+    # 2. Also include already-running devices (GPU-only), otherwise their free_mem will
+    #    always be very low (since their VRAM contains the model).
+    #    These already-running devices probably aren't terrible, since they were picked in the past.
+    #    Worst case, the user can restart the program and that'll get rid of them.
+    devices = list(filter((lambda x: x['mem_free'] > mem_free_threshold or x['device'] in currently_active_devices), devices))
+    devices = list(map(lambda x: x['device'], devices))
+    return devices
+
+def device_init(thread_data, device):
+    '''
+    This function assumes the 'device' has already been verified to be compatible.
+    `get_device_delta()` has already filtered out incompatible devices.
+    '''
+
+    validate_device_id(device, log_prefix='device_init')
+
+    if device == 'cpu':
+        thread_data.device = 'cpu'
+        thread_data.device_name = get_processor_name()
+        print('Render device CPU available as', thread_data.device_name)
+        return
+
+    thread_data.device_name = torch.cuda.get_device_name(device)
+    thread_data.device = device
+
+    # Force full precision on 1660 and 1650 NVIDIA cards to avoid creating green images
+    device_name = thread_data.device_name.lower()
+    thread_data.force_full_precision = ('nvidia' in device_name or 'geforce' in device_name) and (' 1660' in device_name or ' 1650' in device_name)
+    if thread_data.force_full_precision:
+        print('forcing full precision on NVIDIA 16xx cards, to avoid green images. GPU detected: ', thread_data.device_name)
+        # Apply force_full_precision now before models are loaded.
+        thread_data.precision = 'full'
+
+    print(f'Setting {device} as active')
+    torch.cuda.device(device)
+
+    return
+
+def validate_device_id(device, log_prefix=''):
+    def is_valid():
+        if not isinstance(device, str):
+            return False
+        if device == 'cpu':
+            return True
+        if not device.startswith('cuda:') or not device[5:].isnumeric():
+            return False
+        return True
+
+    if not is_valid():
+        raise EnvironmentError(f"{log_prefix}: device id should be 'cpu', or 'cuda:N' (where N is an integer index for the GPU). Got: {device}")
+
+def is_device_compatible(device):
+    '''
+    Returns True/False, and prints any compatibility errors
+    '''
+    try:
+        validate_device_id(device, log_prefix='is_device_compatible')
+    except:
+        print(str(e))
+        return False
+
+    if device == 'cpu': return True
+    # Memory check
+    try:
+        _, mem_total = torch.cuda.mem_get_info(device)
+        mem_total /= float(10**9)
+        if mem_total < 3.0:
+            print(f'GPU {device} with less than 3 GB of VRAM is not compatible with Stable Diffusion')
+            return False
+    except RuntimeError as e:
+        print(str(e))
+        return False
+    return True
+
+def get_processor_name():
+    try:
+        import platform, subprocess
+        if platform.system() == "Windows":
+            return platform.processor()
+        elif platform.system() == "Darwin":
+            os.environ['PATH'] = os.environ['PATH'] + os.pathsep + '/usr/sbin'
+            command = "sysctl -n machdep.cpu.brand_string"
+            return subprocess.check_output(command).strip()
+        elif platform.system() == "Linux":
+            command = "cat /proc/cpuinfo"
+            all_info = subprocess.check_output(command, shell=True).decode().strip()
+            for line in all_info.split("\n"):
+                if "model name" in line:
+                    return re.sub(".*model name.*:", "", line, 1).strip()
+    except:
+        print(traceback.format_exc())
+        return "cpu"
--- a/ui/sd_internal/runtime.py
+++ b/ui/sd_internal/runtime.py
@ -1,8 +1,15 @@
+"""runtime.py: torch device owned by a thread.
+Notes:
+    Avoid device switching, transfering all models will get too complex.
+    To use a diffrent device signal the current render device to exit
+    And then start a new clean thread for the new device.
+"""
 import json
 import os, re
 import traceback
 import torch
 import numpy as np
+from gc import collect as gc_collect
 from omegaconf import OmegaConf
 from PIL import Image, ImageOps
 from tqdm import tqdm, trange
@ -28,70 +35,64 @@ logging.set_verbosity_error()
 # consts
 config_yaml = "optimizedSD/v1-inference.yaml"
 filename_regex = re.compile('[^a-zA-Z0-9]')
+force_gfpgan_to_cuda0 = True # workaround: gfpgan currently works only on cuda:0

 # api stuff
+from sd_internal import device_manager
 from . import Request, Response, Image as ResponseImage
 import base64
 from io import BytesIO
 #from colorama import Fore

-# local
-stop_processing = False
-temp_images = {}
+from threading import local as LocalThreadVars
+thread_data = LocalThreadVars()

-ckpt_file = None
-gfpgan_file = None
-real_esrgan_file = None
+def thread_init(device):
+    # Thread bound properties
+    thread_data.stop_processing = False
+    thread_data.temp_images = {}

-model = None
-modelCS = None
-modelFS = None
-model_gfpgan = None
-model_real_esrgan = None
+    thread_data.ckpt_file = None
+    thread_data.vae_file = None
+    thread_data.gfpgan_file = None
+    thread_data.real_esrgan_file = None

-model_is_half = False
-model_fs_is_half = False
-device = None
-unet_bs = 1
-precision = 'autocast'
-sampler_plms = None
-sampler_ddim = None
+    thread_data.model = None
+    thread_data.modelCS = None
+    thread_data.modelFS = None
+    thread_data.model_gfpgan = None
+    thread_data.model_real_esrgan = None

-has_valid_gpu = False
-force_full_precision = False
-try:
-    gpu = torch.cuda.current_device()
-    gpu_name = torch.cuda.get_device_name(gpu)
-    print('GPU detected: ', gpu_name)
+    thread_data.model_is_half = False
+    thread_data.model_fs_is_half = False
+    thread_data.device = None
+    thread_data.device_name = None
+    thread_data.unet_bs = 1
+    thread_data.precision = 'autocast'
+    thread_data.sampler_plms = None
+    thread_data.sampler_ddim = None

-    force_full_precision = ('nvidia' in gpu_name.lower() or 'geforce' in gpu_name.lower()) and (' 1660' in gpu_name or ' 1650' in gpu_name) # otherwise these NVIDIA cards create green images
-    if force_full_precision:
-        print('forcing full precision on NVIDIA 16xx cards, to avoid green images. GPU detected: ', gpu_name)
+    thread_data.turbo = False
+    thread_data.force_full_precision = False
+    thread_data.reduced_memory = True

-    mem_free, mem_total = torch.cuda.mem_get_info(gpu)
-    mem_total /= float(10**9)
-    if mem_total < 3.0:
-        print("GPUs with less than 3 GB of VRAM are not compatible with Stable Diffusion")
-        raise Exception()
+    device_manager.device_init(thread_data, device)

-    has_valid_gpu = True
-except:
-    print('WARNING: No compatible GPU found. Using the CPU, but this will be very slow!')
-    pass
+def load_model_ckpt():
+    if not thread_data.ckpt_file: raise ValueError(f'Thread ckpt_file is undefined.')
+    if not os.path.exists(thread_data.ckpt_file + '.ckpt'): raise FileNotFoundError(f'Cannot find {thread_data.ckpt_file}.ckpt')

-def load_model_ckpt(ckpt_to_use, device_to_use='cuda', turbo=False, unet_bs_to_use=1, precision_to_use='autocast'):
-    global ckpt_file, model, modelCS, modelFS, model_is_half, device, unet_bs, precision, model_fs_is_half
+    if not thread_data.precision:
+        thread_data.precision = 'full' if thread_data.force_full_precision else 'autocast'

-    device = device_to_use if has_valid_gpu else 'cpu'
-    precision = precision_to_use if not force_full_precision else 'full'
-    unet_bs = unet_bs_to_use
+    if not thread_data.unet_bs:
+        thread_data.unet_bs = 1

-    unload_model()
+    if thread_data.device == 'cpu':
+        thread_data.precision = 'full'

-    if device == 'cpu':
-        precision = 'full'
-
-    sd = load_model_from_config(f"{ckpt_to_use}.ckpt")
+    print('loading', thread_data.ckpt_file + '.ckpt', 'to device', thread_data.device, 'using precision', thread_data.precision)
+    sd = load_model_from_config(thread_data.ckpt_file + '.ckpt')
    li, lo = [], []
    for key, value in sd.items():
        sp = key.split(".")
@ -114,114 +115,205 @@ def load_model_ckpt(ckpt_to_use, device_to_use='cuda', turbo=False, unet_bs_to_u
    model = instantiate_from_config(config.modelUNet)
    _, _ = model.load_state_dict(sd, strict=False)
    model.eval()
-    model.cdevice = device
-    model.unet_bs = unet_bs
-    model.turbo = turbo
+    model.cdevice = torch.device(thread_data.device)
+    model.unet_bs = thread_data.unet_bs
+    model.turbo = thread_data.turbo
+    # if thread_data.device != 'cpu':
+    #     model.to(thread_data.device)
+    #if thread_data.reduced_memory:
+        #model.model1.to("cpu")
+        #model.model2.to("cpu")
+    thread_data.model = model

    modelCS = instantiate_from_config(config.modelCondStage)
    _, _ = modelCS.load_state_dict(sd, strict=False)
    modelCS.eval()
-    modelCS.cond_stage_model.device = device
+    modelCS.cond_stage_model.device = torch.device(thread_data.device)
+    # if thread_data.device != 'cpu':
+    #     if thread_data.reduced_memory:
+    #         modelCS.to('cpu')
+    #     else:
+    #         modelCS.to(thread_data.device) # Preload on device if not already there.
+    thread_data.modelCS = modelCS

    modelFS = instantiate_from_config(config.modelFirstStage)
    _, _ = modelFS.load_state_dict(sd, strict=False)
+
+    if thread_data.vae_file is not None:
+        for model_extension in ['.ckpt', '.vae.pt']:
+            if os.path.exists(thread_data.vae_file + model_extension):
+                print(f"Loading VAE weights from: {thread_data.vae_file}{model_extension}")
+                vae_ckpt = torch.load(thread_data.vae_file + model_extension, map_location="cpu")
+                vae_dict = {k: v for k, v in vae_ckpt["state_dict"].items() if k[0:4] != "loss"}
+                modelFS.first_stage_model.load_state_dict(vae_dict, strict=False)
+                break
+            else:
+                print(f'Cannot find VAE file: {thread_data.vae_file}{model_extension}')
+
    modelFS.eval()
+    # if thread_data.device != 'cpu':
+    #     if thread_data.reduced_memory:
+    #         modelFS.to('cpu')
+    #     else:
+    #         modelFS.to(thread_data.device) # Preload on device if not already there.
+    thread_data.modelFS = modelFS
    del sd

-    if device != "cpu" and precision == "autocast":
-        model.half()
-        modelCS.half()
-        modelFS.half()
-        model_is_half = True
-        model_fs_is_half = True
+    if thread_data.device != "cpu" and thread_data.precision == "autocast":
+        thread_data.model.half()
+        thread_data.modelCS.half()
+        thread_data.modelFS.half()
+        thread_data.model_is_half = True
+        thread_data.model_fs_is_half = True
    else:
-        model_is_half = False
-        model_fs_is_half = False
+        thread_data.model_is_half = False
+        thread_data.model_fs_is_half = False

-    ckpt_file = ckpt_to_use
+    print(f'''loaded model
+ model file: {thread_data.ckpt_file}.ckpt
+ model.device: {model.device}
+ modelCS.device: {modelCS.cond_stage_model.device}
+ modelFS.device: {thread_data.modelFS.device}
+ using precision: {thread_data.precision}''')

-    print('loaded ', ckpt_file, 'to', device, 'precision', precision)
+def unload_filters():
+    if thread_data.model_gfpgan is not None:
+        if thread_data.device != 'cpu': thread_data.model_gfpgan.gfpgan.to('cpu')

-def unload_model():
-    global model, modelCS, modelFS
+        del thread_data.model_gfpgan
+    thread_data.model_gfpgan = None

-    if model is not None:
-        del model
-        del modelCS
-        del modelFS
+    if thread_data.model_real_esrgan is not None:
+        if thread_data.device != 'cpu': thread_data.model_real_esrgan.model.to('cpu')

-    model = None
-    modelCS = None
-    modelFS = None
+        del thread_data.model_real_esrgan
+    thread_data.model_real_esrgan = None

-def load_model_gfpgan(gfpgan_to_use):
-    global gfpgan_file, model_gfpgan
+    gc()

-    if gfpgan_to_use is None:
-        return
+def unload_models():
+    if thread_data.model is not None:
+        print('Unloading models...')
+        if thread_data.device != 'cpu':
+            thread_data.modelFS.to('cpu')
+            thread_data.modelCS.to('cpu')
+            thread_data.model.model1.to("cpu")
+            thread_data.model.model2.to("cpu")

-    gfpgan_file = gfpgan_to_use
-    model_path = gfpgan_to_use + ".pth"
+        del thread_data.model
+        del thread_data.modelCS
+        del thread_data.modelFS

-    if device == 'cpu':
-        model_gfpgan = GFPGANer(model_path=model_path, upscale=1, arch='clean', channel_multiplier=2, bg_upsampler=None, device=torch.device('cpu'))
-    else:
-        model_gfpgan = GFPGANer(model_path=model_path, upscale=1, arch='clean', channel_multiplier=2, bg_upsampler=None, device=torch.device('cuda'))
+    thread_data.model = None
+    thread_data.modelCS = None
+    thread_data.modelFS = None

-    print('loaded ', gfpgan_to_use, 'to', device, 'precision', precision)
+    gc()

-def load_model_real_esrgan(real_esrgan_to_use):
-    global real_esrgan_file, model_real_esrgan
+def wait_model_move_to(model, target_device): # Send to target_device and wait until complete.
+    if thread_data.device == target_device: return
+    start_mem = torch.cuda.memory_allocated(thread_data.device) / 1e6
+    if start_mem <= 0: return
+    model_name = model.__class__.__name__
+    print(f'Device {thread_data.device} - Sending model {model_name} to {target_device} | Memory transfer starting. Memory Used: {round(start_mem)}Mb')
+    start_time = time.time()
+    model.to(target_device)
+    time_step = start_time
+    WARNING_TIMEOUT = 1.5 # seconds - Show activity in console after timeout.
+    last_mem = start_mem
+    is_transfering = True
+    while is_transfering:
+        time.sleep(0.5) # 500ms
+        mem = torch.cuda.memory_allocated(thread_data.device) / 1e6
+        is_transfering = bool(mem > 0 and mem < last_mem) # still stuff loaded, but less than last time.
+        last_mem = mem
+        if not is_transfering:
+            break;
+        if time.time() - time_step > WARNING_TIMEOUT: # Long delay, print to console to show activity.
+            print(f'Device {thread_data.device} - Waiting for Memory transfer. Memory Used: {round(mem)}Mb, Transfered: {round(start_mem - mem)}Mb')
+            time_step = time.time()
+    print(f'Device {thread_data.device} - {model_name} Moved: {round(start_mem - last_mem)}Mb in {round(time.time() - start_time, 3)} seconds to {target_device}')

-    if real_esrgan_to_use is None:
-        return
+def load_model_gfpgan():
+    if thread_data.gfpgan_file is None: raise ValueError(f'Thread gfpgan_file is undefined.')
+    model_path = thread_data.gfpgan_file + ".pth"
+    device = 'cuda:0' if force_gfpgan_to_cuda0 else thread_data.device
+    thread_data.model_gfpgan = GFPGANer(device=torch.device(device), model_path=model_path, upscale=1, arch='clean', channel_multiplier=2, bg_upsampler=None)
+    print('loaded', thread_data.gfpgan_file, 'to', thread_data.model_gfpgan.device, 'precision', thread_data.precision)

-    real_esrgan_file = real_esrgan_to_use
-    model_path = real_esrgan_to_use + ".pth"
+def load_model_real_esrgan():
+    if thread_data.real_esrgan_file is None: raise ValueError(f'Thread real_esrgan_file is undefined.')
+    model_path = thread_data.real_esrgan_file + ".pth"

    RealESRGAN_models = {
        'RealESRGAN_x4plus': RRDBNet(num_in_ch=3, num_out_ch=3, num_feat=64, num_block=23, num_grow_ch=32, scale=4),
        'RealESRGAN_x4plus_anime_6B': RRDBNet(num_in_ch=3, num_out_ch=3, num_feat=64, num_block=6, num_grow_ch=32, scale=4)
    }

-    model_to_use = RealESRGAN_models[real_esrgan_to_use]
+    model_to_use = RealESRGAN_models[thread_data.real_esrgan_file]

-    if device == 'cpu':
-        model_real_esrgan = RealESRGANer(scale=2, model_path=model_path, model=model_to_use, pre_pad=0, half=False) # cpu does not support half
-        model_real_esrgan.device = torch.device('cpu')
-        model_real_esrgan.model.to('cpu')
+    if thread_data.device == 'cpu':
+        thread_data.model_real_esrgan = RealESRGANer(device=torch.device(thread_data.device), scale=2, model_path=model_path, model=model_to_use, pre_pad=0, half=False) # cpu does not support half
+        #thread_data.model_real_esrgan.device = torch.device(thread_data.device)
+        thread_data.model_real_esrgan.model.to('cpu')
    else:
-        model_real_esrgan = RealESRGANer(scale=2, model_path=model_path, model=model_to_use, pre_pad=0, half=model_is_half)
+        thread_data.model_real_esrgan = RealESRGANer(device=torch.device(thread_data.device), scale=2, model_path=model_path, model=model_to_use, pre_pad=0, half=thread_data.model_is_half)

-    model_real_esrgan.model.name = real_esrgan_to_use
+    thread_data.model_real_esrgan.model.name = thread_data.real_esrgan_file
+    print('loaded ', thread_data.real_esrgan_file, 'to', thread_data.model_real_esrgan.device, 'precision', thread_data.precision)

-    print('loaded ', real_esrgan_to_use, 'to', device, 'precision', precision)
+
+def get_session_out_path(disk_path, session_id):
+    if disk_path is None: return None
+    if session_id is None: return None
+
+    session_out_path = os.path.join(disk_path, filename_regex.sub('_',session_id))
+    os.makedirs(session_out_path, exist_ok=True)
+    return session_out_path

 def get_base_path(disk_path, session_id, prompt, img_id, ext, suffix=None):
    if disk_path is None: return None
    if session_id is None: return None
    if ext is None: raise Exception('Missing ext')

-    session_out_path = os.path.join(disk_path, session_id)
-    os.makedirs(session_out_path, exist_ok=True)
+    session_out_path = get_session_out_path(disk_path, session_id)

    prompt_flattened = filename_regex.sub('_', prompt)[:50]
-    

    if suffix is not None:
        return os.path.join(session_out_path, f"{prompt_flattened}_{img_id}_{suffix}.{ext}")
    return os.path.join(session_out_path, f"{prompt_flattened}_{img_id}.{ext}")

-def apply_filters(filter_name, image_data):
+def apply_filters(filter_name, image_data, model_path=None):
    print(f'Applying filter {filter_name}...')
-    gc()
+    gc() # Free space before loading new data.

    if filter_name == 'gfpgan':
-        _, _, output = model_gfpgan.enhance(image_data[:,:,::-1], has_aligned=False, only_center_face=False, paste_back=True)
+        if isinstance(image_data, torch.Tensor):
+            image_data.to('cuda:0' if force_gfpgan_to_cuda0 else thread_data.device)
+
+        if model_path is not None and model_path != thread_data.gfpgan_file:
+            thread_data.gfpgan_file = model_path
+            load_model_gfpgan()
+        elif not thread_data.model_gfpgan:
+            load_model_gfpgan()
+        if thread_data.model_gfpgan is None: raise Exception('Model "gfpgan" not loaded.')
+        print('enhance with', thread_data.gfpgan_file, 'on', thread_data.model_gfpgan.device, 'precision', thread_data.precision)
+        _, _, output = thread_data.model_gfpgan.enhance(image_data[:,:,::-1], has_aligned=False, only_center_face=False, paste_back=True)
        image_data = output[:,:,::-1]

    if filter_name == 'real_esrgan':
-        output, _ = model_real_esrgan.enhance(image_data[:,:,::-1])
+        if isinstance(image_data, torch.Tensor):
+            image_data.to(thread_data.device)
+
+        if model_path is not None and model_path != thread_data.real_esrgan_file:
+            thread_data.real_esrgan_file = model_path
+            load_model_real_esrgan()
+        elif not thread_data.model_real_esrgan:
+            load_model_real_esrgan()
+        if thread_data.model_real_esrgan is None: raise Exception('Model "gfpgan" not loaded.')
+        print('enhance with', thread_data.real_esrgan_file, 'on', thread_data.model_real_esrgan.device, 'precision', thread_data.precision)
+        output, _ = thread_data.model_real_esrgan.enhance(image_data[:,:,::-1])
        image_data = output[:,:,::-1]

    return image_data
@ -232,83 +324,102 @@ def mk_img(req: Request):
    except Exception as e:
        print(traceback.format_exc())

-        gc()
-
-        if device != "cpu":
-            modelFS.to("cpu")
-            modelCS.to("cpu")
-
-            model.model1.to("cpu")
-            model.model2.to("cpu")
-
-        gc()
+        if thread_data.device != 'cpu':
+            thread_data.modelFS.to('cpu')
+            thread_data.modelCS.to('cpu')
+            thread_data.model.model1.to("cpu")
+            thread_data.model.model2.to("cpu")

+        gc() # Release from memory.
        yield json.dumps({
            "status": 'failed',
            "detail": str(e)
        })

-def do_mk_img(req: Request):
-    global ckpt_file
-    global model, modelCS, modelFS, device
-    global model_gfpgan, model_real_esrgan
-    global stop_processing
+def update_temp_img(req, x_samples):
+    partial_images = []
+    for i in range(req.num_outputs):
+        x_sample_ddim = thread_data.modelFS.decode_first_stage(x_samples[i].unsqueeze(0))
+        x_sample = torch.clamp((x_sample_ddim + 1.0) / 2.0, min=0.0, max=1.0)
+        x_sample = 255.0 * rearrange(x_sample[0].cpu().numpy(), "c h w -> h w c")
+        x_sample = x_sample.astype(np.uint8)
+        img = Image.fromarray(x_sample)
+        buf = BytesIO()
+        img.save(buf, format='JPEG')
+        buf.seek(0)

-    stop_processing = False
+        del img, x_sample, x_sample_ddim
+        # don't delete x_samples, it is used in the code that called this callback
+
+        thread_data.temp_images[str(req.session_id) + '/' + str(i)] = buf
+        partial_images.append({'path': f'/image/tmp/{req.session_id}/{i}'})
+    return partial_images
+
+# Build and return the apropriate generator for do_mk_img
+def get_image_progress_generator(req, extra_props=None):
+    if not req.stream_progress_updates:
+        def empty_callback(x_samples, i): return x_samples
+        return empty_callback
+
+    thread_data.partial_x_samples = None
+    last_callback_time = -1
+    def img_callback(x_samples, i):
+        nonlocal last_callback_time
+
+        thread_data.partial_x_samples = x_samples
+        step_time = time.time() - last_callback_time if last_callback_time != -1 else -1
+        last_callback_time = time.time()
+
+        progress = {"step": i, "step_time": step_time}
+        if extra_props is not None:
+            progress.update(extra_props)
+
+        if req.stream_image_progress and i % 5 == 0:
+            progress['output'] = update_temp_img(req, x_samples)
+
+        yield json.dumps(progress)
+
+        if thread_data.stop_processing:
+            raise UserInitiatedStop("User requested that we stop processing")
+    return img_callback
+
+def do_mk_img(req: Request):
+    thread_data.stop_processing = False

    res = Response()
    res.request = req
    res.images = []

-    temp_images.clear()
+    thread_data.temp_images.clear()

    # custom model support:
    #  the req.use_stable_diffusion_model needs to be a valid path
    #  to the ckpt file (without the extension).
+    if not os.path.exists(req.use_stable_diffusion_model + '.ckpt'): raise FileNotFoundError(f'Cannot find {req.use_stable_diffusion_model}.ckpt')

    needs_model_reload = False
-    ckpt_to_use = ckpt_file
-    if ckpt_to_use != req.use_stable_diffusion_model:
-        ckpt_to_use = req.use_stable_diffusion_model
+    if not thread_data.model or thread_data.ckpt_file != req.use_stable_diffusion_model or thread_data.vae_file != req.use_vae_model:
+        thread_data.ckpt_file = req.use_stable_diffusion_model
+        thread_data.vae_file = req.use_vae_model
        needs_model_reload = True

-    model.turbo = req.turbo
-    if req.use_cpu:
-        if device != 'cpu':
-            device = 'cpu'
-
-            if model_is_half:
-                load_model_ckpt(ckpt_to_use, device)
-                needs_model_reload = False
-
-            load_model_gfpgan(gfpgan_file)
-            load_model_real_esrgan(real_esrgan_file)
-    else:
-        if has_valid_gpu:
-            prev_device = device
-            device = 'cuda'
-
-            if (precision == 'autocast' and (req.use_full_precision or not model_is_half)) or \
-                (precision == 'full' and not req.use_full_precision and not force_full_precision):
-
-                load_model_ckpt(ckpt_to_use, device, req.turbo, unet_bs, ('full' if req.use_full_precision else 'autocast'))
-                needs_model_reload = False
-
-                if prev_device != device:
-                    load_model_gfpgan(gfpgan_file)
-                    load_model_real_esrgan(real_esrgan_file)
+    if thread_data.device != 'cpu':
+        if (thread_data.precision == 'autocast' and (req.use_full_precision or not thread_data.model_is_half)) or \
+            (thread_data.precision == 'full' and not req.use_full_precision and not thread_data.force_full_precision):
+            thread_data.precision = 'full' if req.use_full_precision else 'autocast'
+            needs_model_reload = True

    if needs_model_reload:
-        load_model_ckpt(ckpt_to_use, device, req.turbo, unet_bs, precision)
+        unload_models()
+        unload_filters()
+        load_model_ckpt()

-    if req.use_face_correction != gfpgan_file:
-        load_model_gfpgan(req.use_face_correction)
+    if thread_data.turbo != req.turbo:
+        thread_data.turbo = req.turbo
+        thread_data.model.turbo = req.turbo

-    if req.use_upscale != real_esrgan_file:
-        load_model_real_esrgan(req.use_upscale)
-
-    model.cdevice = device
-    modelCS.cond_stage_model.device = device
+    # Start by cleaning memory, loading and unloading things can leave memory allocated.
+    gc()

    opt_prompt = req.prompt
    opt_seed = req.seed
@ -316,11 +427,9 @@ def do_mk_img(req: Request):
    opt_C = 4
    opt_f = 8
    opt_ddim_eta = 0.0
-    opt_init_img = req.init_image

-    print(req.to_string(), '\n    device', device)
-
-    print('\n\n    Using precision:', precision)
+    print(req, '\n    device', torch.device(thread_data.device), "as", thread_data.device_name)
+    print('\n\n    Using precision:', thread_data.precision)

    seed_everything(opt_seed)

@ -329,7 +438,7 @@ def do_mk_img(req: Request):
    assert prompt is not None
    data = [batch_size * [prompt]]

-    if precision == "autocast" and device != "cpu":
+    if thread_data.precision == "autocast" and thread_data.device != "cpu":
        precision_scope = autocast
    else:
        precision_scope = nullcontext
@ -345,46 +454,46 @@ def do_mk_img(req: Request):
        handler = _img2img

        init_image = load_img(req.init_image, req.width, req.height)
-        init_image = init_image.to(device)
+        init_image = init_image.to(thread_data.device)

-        if device != "cpu" and precision == "autocast":
+        if thread_data.device != "cpu" and thread_data.precision == "autocast":
            init_image = init_image.half()

-        modelFS.to(device)
+        thread_data.modelFS.to(thread_data.device)

        init_image = repeat(init_image, '1 ... -> b ...', b=batch_size)
-        init_latent = modelFS.get_first_stage_encoding(modelFS.encode_first_stage(init_image))  # move to latent space
+        init_latent = thread_data.modelFS.get_first_stage_encoding(thread_data.modelFS.encode_first_stage(init_image))  # move to latent space

        if req.mask is not None:
-            mask = load_mask(req.mask, req.width, req.height, init_latent.shape[2], init_latent.shape[3], True).to(device)
+            mask = load_mask(req.mask, req.width, req.height, init_latent.shape[2], init_latent.shape[3], True).to(thread_data.device)
            mask = mask[0][0].unsqueeze(0).repeat(4, 1, 1).unsqueeze(0)
            mask = repeat(mask, '1 ... -> b ...', b=batch_size)

-            if device != "cpu" and precision == "autocast":
+            if thread_data.device != "cpu" and thread_data.precision == "autocast":
                mask = mask.half()

-        move_fs_to_cpu()
+        # Send to CPU and wait until complete.
+        wait_model_move_to(thread_data.modelFS, 'cpu')

        assert 0. <= req.prompt_strength <= 1., 'can only work with strength in [0.0, 1.0]'
        t_enc = int(req.prompt_strength * req.num_inference_steps)
        print(f"target t_enc is {t_enc} steps")

    if req.save_to_disk_path is not None:
-        session_out_path = os.path.join(req.save_to_disk_path, req.session_id)
-        os.makedirs(session_out_path, exist_ok=True)
+        session_out_path = get_session_out_path(req.save_to_disk_path, req.session_id)
    else:
        session_out_path = None

-    seeds = ""
    with torch.no_grad():
        for n in trange(opt_n_iter, desc="Sampling"):
            for prompts in tqdm(data, desc="data"):

                with precision_scope("cuda"):
-                    modelCS.to(device)
+                    if thread_data.reduced_memory:
+                        thread_data.modelCS.to(thread_data.device)
                    uc = None
                    if req.guidance_scale != 1.0:
-                        uc = modelCS.get_learned_conditioning(batch_size * [req.negative_prompt])
+                        uc = thread_data.modelCS.get_learned_conditioning(batch_size * [req.negative_prompt])
                    if isinstance(prompts, tuple):
                        prompts = list(prompts)

@ -397,85 +506,65 @@ def do_mk_img(req: Request):
                            weight = weights[i]
                            # if not skip_normalize:
                            weight = weight / totalWeight
-                            c = torch.add(c, modelCS.get_learned_conditioning(subprompts[i]), alpha=weight)
+                            c = torch.add(c, thread_data.modelCS.get_learned_conditioning(subprompts[i]), alpha=weight)
                    else:
-                        c = modelCS.get_learned_conditioning(prompts)
+                        c = thread_data.modelCS.get_learned_conditioning(prompts)

-                    modelFS.to(device)
+                    if thread_data.reduced_memory:
+                        thread_data.modelFS.to(thread_data.device)

-                    partial_x_samples = None
-                    last_callback_time = -1
-                    def img_callback(x_samples, i):
-                        nonlocal partial_x_samples, last_callback_time
-
-                        partial_x_samples = x_samples
-
-                        if req.stream_progress_updates:
-                            n_steps = req.num_inference_steps if req.init_image is None else t_enc
-                            step_time = time.time() - last_callback_time if last_callback_time != -1 else -1
-                            last_callback_time = time.time()
-
-                            progress = {"step": i, "total_steps": n_steps, "step_time": step_time}
-
-                            if req.stream_image_progress and i % 5 == 0:
-                                partial_images = []
-
-                                for i in range(batch_size):
-                                    x_samples_ddim = modelFS.decode_first_stage(x_samples[i].unsqueeze(0))
-                                    x_sample = torch.clamp((x_samples_ddim + 1.0) / 2.0, min=0.0, max=1.0)
-                                    x_sample = 255.0 * rearrange(x_sample[0].cpu().numpy(), "c h w -> h w c")
-                                    x_sample = x_sample.astype(np.uint8)
-                                    img = Image.fromarray(x_sample)
-                                    buf = BytesIO()
-                                    img.save(buf, format='JPEG')
-                                    buf.seek(0)
-
-                                    del img, x_sample, x_samples_ddim
-                                    # don't delete x_samples, it is used in the code that called this callback
-
-                                    temp_images[str(req.session_id) + '/' + str(i)] = buf
-                                    partial_images.append({'path': f'/image/tmp/{req.session_id}/{i}'})
-
-                                progress['output'] = partial_images
-
-                            yield json.dumps(progress)
-
-                        if stop_processing:
-                            raise UserInitiatedStop("User requested that we stop processing")
+                    n_steps = req.num_inference_steps if req.init_image is None else t_enc
+                    img_callback = get_image_progress_generator(req, {"total_steps": n_steps})

                    # run the handler
                    try:
+                        print('Running handler...')
                        if handler == _txt2img:
                            x_samples = _txt2img(req.width, req.height, req.num_outputs, req.num_inference_steps, req.guidance_scale, None, opt_C, opt_f, opt_ddim_eta, c, uc, opt_seed, img_callback, mask, req.sampler)
                        else:
                            x_samples = _img2img(init_latent, t_enc, batch_size, req.guidance_scale, c, uc, req.num_inference_steps, opt_ddim_eta, opt_seed, img_callback, mask)

-                        yield from x_samples
-
-                        x_samples = partial_x_samples
+                        if req.stream_progress_updates:
+                            yield from x_samples
+                        if hasattr(thread_data, 'partial_x_samples'):
+                            if thread_data.partial_x_samples is not None:
+                                x_samples = thread_data.partial_x_samples
+                            del thread_data.partial_x_samples
                    except UserInitiatedStop:
-                        if partial_x_samples is None:
+                        if not hasattr(thread_data, 'partial_x_samples'):
                            continue
+                        if thread_data.partial_x_samples is None:
+                            del thread_data.partial_x_samples
+                            continue
+                        x_samples = thread_data.partial_x_samples
+                        del thread_data.partial_x_samples

-                        x_samples = partial_x_samples
-
-                    print("saving images")
+                    print("decoding images")
+                    img_data = [None] * batch_size
                    for i in range(batch_size):
-                        img_id = base64.b64encode(int(time.time()+i).to_bytes(8, 'big')).decode() # Generate unique ID based on time.
-                        img_id = img_id.translate({43:None, 47:None, 61:None})[-8:] # Remove + / = and keep last 8 chars.
-
-                        x_samples_ddim = modelFS.decode_first_stage(x_samples[i].unsqueeze(0))
+                        x_samples_ddim = thread_data.modelFS.decode_first_stage(x_samples[i].unsqueeze(0))
                        x_sample = torch.clamp((x_samples_ddim + 1.0) / 2.0, min=0.0, max=1.0)
                        x_sample = 255.0 * rearrange(x_sample[0].cpu().numpy(), "c h w -> h w c")
                        x_sample = x_sample.astype(np.uint8)
-                        img = Image.fromarray(x_sample)
+                        img_data[i] = x_sample
+                    del x_samples, x_samples_ddim, x_sample
+
+                    if thread_data.reduced_memory:
+                        # Send to CPU and wait until complete.
+                        wait_model_move_to(thread_data.modelFS, 'cpu')
+
+                    print("saving images")
+                    for i in range(batch_size):
+                        img = Image.fromarray(img_data[i])
+                        img_id = base64.b64encode(int(time.time()+i).to_bytes(8, 'big')).decode() # Generate unique ID based on time.
+                        img_id = img_id.translate({43:None, 47:None, 61:None})[-8:] # Remove + / = and keep last 8 chars.

                        has_filters =   (req.use_face_correction is not None and req.use_face_correction.startswith('GFPGAN')) or \
                                        (req.use_upscale is not None and req.use_upscale.startswith('RealESRGAN'))

                        return_orig_img = not has_filters or not req.show_only_filtered_image

-                        if stop_processing:
+                        if thread_data.stop_processing:
                            return_orig_img = True

                        if req.save_to_disk_path is not None:
@ -486,25 +575,24 @@ def do_mk_img(req: Request):
                            save_metadata(meta_out_path, req, prompts[0], opt_seed)

                        if return_orig_img:
-                            img_data = img_to_base64_str(img, req.output_format)
-                            res_image_orig = ResponseImage(data=img_data, seed=opt_seed)
+                            img_str = img_to_base64_str(img, req.output_format)
+                            res_image_orig = ResponseImage(data=img_str, seed=opt_seed)
                            res.images.append(res_image_orig)

                            if req.save_to_disk_path is not None:
                                res_image_orig.path_abs = img_out_path
-
                        del img

-                        if has_filters and not stop_processing:
+                        if has_filters and not thread_data.stop_processing:
                            filters_applied = []
                            if req.use_face_correction:
-                                x_sample = apply_filters('gfpgan', x_sample)
+                                img_data[i] = apply_filters('gfpgan', img_data[i], req.use_face_correction)
                                filters_applied.append(req.use_face_correction)
                            if req.use_upscale:
-                                x_sample = apply_filters('real_esrgan', x_sample)
+                                img_data[i] = apply_filters('real_esrgan', img_data[i], req.use_upscale)
                                filters_applied.append(req.use_upscale)
                            if (len(filters_applied) > 0):
-                                filtered_image = Image.fromarray(x_sample)
+                                filtered_image = Image.fromarray(img_data[i])
                                filtered_img_data = img_to_base64_str(filtered_image, req.output_format)
                                response_image = ResponseImage(data=filtered_img_data, seed=opt_seed)
                                res.images.append(response_image)
@ -513,17 +601,17 @@ def do_mk_img(req: Request):
                                    save_image(filtered_image, filtered_img_out_path)
                                    response_image.path_abs = filtered_img_out_path
                                del filtered_image
-
-                        seeds += str(opt_seed) + ","
+                        # Filter Applied, move to next seed
                        opt_seed += 1

-                    move_fs_to_cpu()
+                    # if thread_data.reduced_memory:
+                    #     unload_filters()
+                    del img_data
                    gc()
-                    del x_samples, x_samples_ddim, x_sample
-                    print("memory_final = ", torch.cuda.memory_allocated() / 1e6)
+                    if thread_data.device != 'cpu':
+                        print(f'memory_final = {round(torch.cuda.memory_allocated(thread_data.device) / 1e6, 2)}Mb')

    print('Task completed')
-
    yield json.dumps(res.json())

 def save_image(img, img_out_path):
@ -533,7 +621,7 @@ def save_image(img, img_out_path):
        print('could not save the file', traceback.format_exc())

 def save_metadata(meta_out_path, req, prompt, opt_seed):
-    metadata = f"""{prompt}
+    metadata = f'''{prompt}
 Width: {req.width}
 Height: {req.height}
 Seed: {opt_seed}
@ -544,8 +632,9 @@ Use Face Correction: {req.use_face_correction}
 Use Upscaling: {req.use_upscale}
 Sampler: {req.sampler}
 Negative Prompt: {req.negative_prompt}
-Stable Diffusion Model: {req.use_stable_diffusion_model + '.ckpt'}
-"""
+Stable Diffusion model: {req.use_stable_diffusion_model + '.ckpt'}
+VAE model: {req.use_vae_model}
+'''
    try:
        with open(meta_out_path, 'w', encoding='utf-8') as f:
            f.write(metadata)
@ -555,16 +644,13 @@ Stable Diffusion Model: {req.use_stable_diffusion_model + '.ckpt'}
 def _txt2img(opt_W, opt_H, opt_n_samples, opt_ddim_steps, opt_scale, start_code, opt_C, opt_f, opt_ddim_eta, c, uc, opt_seed, img_callback, mask, sampler_name):
    shape = [opt_n_samples, opt_C, opt_H // opt_f, opt_W // opt_f]

-    if device != "cpu":
-        mem = torch.cuda.memory_allocated() / 1e6
-        modelCS.to("cpu")
-        while torch.cuda.memory_allocated() / 1e6 >= mem:
-            time.sleep(1)
+    # Send to CPU and wait until complete.
+    wait_model_move_to(thread_data.modelCS, 'cpu')

    if sampler_name == 'ddim':
-        model.make_schedule(ddim_num_steps=opt_ddim_steps, ddim_eta=opt_ddim_eta, verbose=False)
+        thread_data.model.make_schedule(ddim_num_steps=opt_ddim_steps, ddim_eta=opt_ddim_eta, verbose=False)

-    samples_ddim = model.sample(
+    samples_ddim = thread_data.model.sample(
        S=opt_ddim_steps,
        conditioning=c,
        seed=opt_seed,
@ -578,14 +664,13 @@ def _txt2img(opt_W, opt_H, opt_n_samples, opt_ddim_steps, opt_scale, start_code,
        mask=mask,
        sampler = sampler_name,
    )
-
    yield from samples_ddim

 def _img2img(init_latent, t_enc, batch_size, opt_scale, c, uc, opt_ddim_steps, opt_ddim_eta, opt_seed, img_callback, mask):
    # encode (scaled latent)
-    z_enc = model.stochastic_encode(
+    z_enc = thread_data.model.stochastic_encode(
        init_latent,
-        torch.tensor([t_enc] * batch_size).to(device),
+        torch.tensor([t_enc] * batch_size).to(thread_data.device),
        opt_seed,
        opt_ddim_eta,
        opt_ddim_steps,
@ -593,7 +678,7 @@ def _img2img(init_latent, t_enc, batch_size, opt_scale, c, uc, opt_ddim_steps, o
    x_T = None if mask is None else init_latent

    # decode it
-    samples_ddim = model.sample(
+    samples_ddim = thread_data.model.sample(
        t_enc,
        c,
        z_enc,
@ -604,20 +689,12 @@ def _img2img(init_latent, t_enc, batch_size, opt_scale, c, uc, opt_ddim_steps, o
        x_T=x_T,
        sampler = 'ddim'
    )
-
    yield from samples_ddim

-def move_fs_to_cpu():
-    if device != "cpu":
-        mem = torch.cuda.memory_allocated() / 1e6
-        modelFS.to("cpu")
-        while torch.cuda.memory_allocated() / 1e6 >= mem:
-            time.sleep(1)
-
 def gc():
-    if device == 'cpu':
+    gc_collect()
+    if thread_data.device == 'cpu':
        return
-
    torch.cuda.empty_cache()
    torch.cuda.ipc_collect()

@ -627,7 +704,6 @@ def chunk(it, size):
    it = iter(it)
    return iter(lambda: tuple(islice(it, size)), ())

-
 def load_model_from_config(ckpt, verbose=False):
    print(f"Loading model from {ckpt}")
    pl_sd = torch.load(ckpt, map_location="cpu")
@ -689,10 +765,14 @@ def img_to_base64_str(img, output_format="PNG"):
    img_str = f"data:{mime_type};base64," + base64.b64encode(img_byte).decode()
    return img_str

-def base64_str_to_img(img_str):
+def base64_str_to_buffer(img_str):
    mime_type = "image/png" if img_str.startswith("data:image/png;") else "image/jpeg"
    img_str = img_str[len(f"data:{mime_type};base64,"):]
    data = base64.b64decode(img_str)
    buffered = BytesIO(data)
+    return buffered
+
+def base64_str_to_img(img_str):
+    buffered = base64_str_to_buffer(img_str)
    img = Image.open(buffered)
    return img
--- a/ui/sd_internal/task_manager.py
+++ b/ui/sd_internal/task_manager.py
@ -1,13 +1,27 @@
+"""task_manager.py: manage tasks dispatching and render threads.
+Notes:
+    render_threads should be the only hard reference held by the manager to the threads.
+    Use weak_thread_data to store all other data using weak keys.
+    This will allow for garbage collection after the thread dies.
+"""
 import json
 import traceback

 TASK_TTL = 15 * 60 # seconds, Discard last session's task timeout

-import queue, threading, time
+import torch
+import queue, threading, time, weakref
 from typing import Any, Generator, Hashable, Optional, Union

 from pydantic import BaseModel
-from sd_internal import Request, Response
+from sd_internal import Request, Response, runtime, device_manager
+
+THREAD_NAME_PREFIX = 'Runtime-Render/'
+ERR_LOCK_FAILED = ' failed to acquire lock within timeout.'
+LOCK_TIMEOUT = 15 # Maximum locking time in seconds before failing a task.
+# It's better to get an exception than a deadlock... ALWAYS use timeout in critical paths.
+
+DEVICE_START_TIMEOUT = 60 # seconds - Maximum time to wait for a render device to init.

 class SymbolClass(type): # Print nicely formatted Symbol names.
    def __repr__(self): return self.__qualname__
@ -25,7 +39,8 @@ class RenderTask(): # Task with output queue and completion lock.
    def __init__(self, req: Request):
        self.request: Request = req # Initial Request
        self.response: Any = None # Copy of the last reponse
-        self.temp_images:[] = [None] * req.num_outputs * (1 if req.show_only_filtered_image else 2)
+        self.render_device = None # Select the task affinity. (Not used to change active devices).
+        self.temp_images:list = [None] * req.num_outputs * (1 if req.show_only_filtered_image else 2)
        self.error: Exception = None
        self.lock: threading.Lock = threading.Lock() # Locks at task start and unlocks when task is completed
        self.buffer_queue: queue.Queue = queue.Queue() # Queue of JSON string segments
@ -55,28 +70,43 @@ class ImageRequest(BaseModel):
    # allow_nsfw: bool = False
    save_to_disk_path: str = None
    turbo: bool = True
-    use_cpu: bool = False
+    use_cpu: bool = False ##TODO Remove after UI and plugins transition.
+    render_device: str = None # Select the task affinity. (Not used to change active devices).
    use_full_precision: bool = False
    use_face_correction: str = None # or "GFPGANv1.3"
    use_upscale: str = None # or "RealESRGAN_x4plus" or "RealESRGAN_x4plus_anime_6B"
    use_stable_diffusion_model: str = "sd-v1-4"
+    use_vae_model: str = None
    show_only_filtered_image: bool = False
    output_format: str = "jpeg" # or "png"

    stream_progress_updates: bool = False
    stream_image_progress: bool = False

+class FilterRequest(BaseModel):
+    session_id: str = "session"
+    model: str = None
+    name: str = ""
+    init_image: str = None # base64
+    width: int = 512
+    height: int = 512
+    save_to_disk_path: str = None
+    turbo: bool = True
+    render_device: str = None
+    use_full_precision: bool = False
+    output_format: str = "jpeg" # or "png"
+
 # Temporary cache to allow to query tasks results for a short time after they are completed.
 class TaskCache():
    def __init__(self):
        self._base = dict()
-        self._lock: threading.Lock = threading.RLock()
+        self._lock: threading.Lock = threading.Lock()
    def _get_ttl_time(self, ttl: int) -> int:
        return int(time.time()) + ttl
    def _is_expired(self, timestamp: int) -> bool:
        return int(time.time()) >= timestamp
    def clean(self) -> None:
-        if not self._lock.acquire(blocking=True, timeout=10): raise Exception('TaskCache.clean failed to acquire lock within timeout.')
+        if not self._lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('TaskCache.clean' + ERR_LOCK_FAILED)
        try:
            # Create a list of expired keys to delete
            to_delete = []
@ -91,11 +121,11 @@ class TaskCache():
        finally:
            self._lock.release()
    def clear(self) -> None:
-        if not self._lock.acquire(blocking=True, timeout=10): raise Exception('TaskCache.clear failed to acquire lock within timeout.')
+        if not self._lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('TaskCache.clear' + ERR_LOCK_FAILED)
        try: self._base.clear()
        finally: self._lock.release()
    def delete(self, key: Hashable) -> bool:
-        if not self._lock.acquire(blocking=True, timeout=10): raise Exception('TaskCache.delete failed to acquire lock within timeout.')
+        if not self._lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('TaskCache.delete' + ERR_LOCK_FAILED)
        try:
            if key not in self._base:
                return False
@ -104,7 +134,7 @@ class TaskCache():
        finally:
            self._lock.release()
    def keep(self, key: Hashable, ttl: int) -> bool:
-        if not self._lock.acquire(blocking=True, timeout=10): raise Exception('TaskCache.keep failed to acquire lock within timeout.')
+        if not self._lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('TaskCache.keep' + ERR_LOCK_FAILED)
        try:
            if key in self._base:
                _, value = self._base.get(key)
@ -114,7 +144,7 @@ class TaskCache():
        finally:
            self._lock.release()
    def put(self, key: Hashable, value: Any, ttl: int) -> bool:
-        if not self._lock.acquire(blocking=True, timeout=10): raise Exception('TaskCache.put failed to acquire lock within timeout.')
+        if not self._lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('TaskCache.put' + ERR_LOCK_FAILED)
        try:
            self._base[key] = (
                self._get_ttl_time(ttl), value
@ -128,131 +158,343 @@ class TaskCache():
        finally:
            self._lock.release()
    def tryGet(self, key: Hashable) -> Any:
-        if not self._lock.acquire(blocking=True, timeout=10): raise Exception('TaskCache.tryGet failed to acquire lock within timeout.')
+        if not self._lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('TaskCache.tryGet' + ERR_LOCK_FAILED)
        try:
            ttl, value = self._base.get(key, (None, None))
            if ttl is not None and self._is_expired(ttl):
                print(f'Session {key} expired. Discarding data.')
-                self.delete(key)
+                del self._base[key]
                return None
            return value
        finally:
            self._lock.release()

+manager_lock = threading.RLock()
+render_threads = []
 current_state = ServerStates.Init
 current_state_error:Exception = None
 current_model_path = None
-tasks_queue = queue.Queue()
+current_vae_path = None
+tasks_queue = []
 task_cache = TaskCache()
 default_model_to_load = None
+default_vae_to_load = None
+weak_thread_data = weakref.WeakKeyDictionary()

-def preload_model(file_path=None):
-    global current_state, current_state_error, current_model_path
-    if file_path == None:
-        file_path = default_model_to_load
-    if file_path == current_model_path:
+def preload_model(ckpt_file_path=None, vae_file_path=None):
+    global current_state, current_state_error, current_model_path, current_vae_path
+    if ckpt_file_path == None:
+        ckpt_file_path = default_model_to_load
+    if vae_file_path == None:
+        vae_file_path = default_vae_to_load
+    if ckpt_file_path == current_model_path and vae_file_path == current_vae_path:
        return
    current_state = ServerStates.LoadingModel
    try:
        from . import runtime
-        runtime.load_model_ckpt(ckpt_to_use=file_path)
-        current_model_path = file_path
+        runtime.thread_data.ckpt_file = ckpt_file_path
+        runtime.thread_data.vae_file = vae_file_path
+        runtime.load_model_ckpt()
+        current_model_path = ckpt_file_path
+        current_vae_path = vae_file_path
        current_state_error = None
        current_state = ServerStates.Online
    except Exception as e:
        current_model_path = None
+        current_vae_path = None
        current_state_error = e
        current_state = ServerStates.Unavailable
        print(traceback.format_exc())

-def thread_render():
-    global current_state, current_state_error, current_model_path
+def thread_get_next_task():
    from . import runtime
-    current_state = ServerStates.Online
-    preload_model()
+    if not manager_lock.acquire(blocking=True, timeout=LOCK_TIMEOUT):
+        print('Render thread on device', runtime.thread_data.device, 'failed to acquire manager lock.')
+        return None
+    if len(tasks_queue) <= 0:
+        manager_lock.release()
+        return None
+    task = None
+    try:  # Select a render task.
+        for queued_task in tasks_queue:
+            if queued_task.request.use_face_correction and runtime.thread_data.device == 'cpu' and is_alive() == 1:
+                queued_task.error = Exception('The CPU cannot be used to run this task currently. Please remove "Fix incorrect faces" from Image Settings and try again.')
+                task = queued_task
+                break
+            if queued_task.render_device and runtime.thread_data.device != queued_task.render_device:
+                # Is asking for a specific render device.
+                if is_alive(queued_task.render_device) > 0:
+                    continue  # requested device alive, skip current one.
+                else:
+                    # Requested device is not active, return error to UI.
+                    queued_task.error = Exception(queued_task.render_device + ' is not currently active.')
+                    task = queued_task
+                    break
+            if not queued_task.render_device and runtime.thread_data.device == 'cpu' and is_alive() > 1:
+                # not asking for any specific devices, cpu want to grab task but other render devices are alive.
+                continue  # Skip Tasks, don't run on CPU unless there is nothing else or user asked for it.
+            task = queued_task
+            break
+        if task is not None:
+            del tasks_queue[tasks_queue.index(task)]
+        return task
+    finally:
+        manager_lock.release()
+
+def thread_render(device):
+    global current_state, current_state_error, current_model_path, current_vae_path
+    from . import runtime
+    try:
+        runtime.thread_init(device)
+    except Exception as e:
+        print(traceback.format_exc())
+        weak_thread_data[threading.current_thread()] = {
+            'error': e
+        }
+        return
+    weak_thread_data[threading.current_thread()] = {
+        'device': runtime.thread_data.device,
+        'device_name': runtime.thread_data.device_name,
+        'alive': True
+    }
+    if runtime.thread_data.device != 'cpu' or is_alive() == 1:
+        preload_model()
+        current_state = ServerStates.Online
    while True:
        task_cache.clean()
+        if not weak_thread_data[threading.current_thread()]['alive']:
+            print(f'Shutting down thread for device {runtime.thread_data.device}')
+            runtime.unload_models()
+            runtime.unload_filters()
+            return
        if isinstance(current_state_error, SystemExit):
            current_state = ServerStates.Unavailable
            return
-        task = None
-        try:
-            task = tasks_queue.get(timeout=1)
-        except queue.Empty as e:
-            if isinstance(current_state_error, SystemExit):
-                current_state = ServerStates.Unavailable
-                return
-            else: continue
-        #if current_model_path != task.request.use_stable_diffusion_model:
-        #    preload_model(task.request.use_stable_diffusion_model)
+        task = thread_get_next_task()
+        if task is None:
+            time.sleep(1)
+            continue
+        if task.error is not None:
+            print(task.error)
+            task.response = {"status": 'failed', "detail": str(task.error)}
+            task.buffer_queue.put(json.dumps(task.response))
+            continue
        if current_state_error:
            task.error = current_state_error
+            task.response = {"status": 'failed', "detail": str(task.error)}
+            task.buffer_queue.put(json.dumps(task.response))
            continue
-        print(f'Session {task.request.session_id} starting task {id(task)}')
+        print(f'Session {task.request.session_id} starting task {id(task)} on {runtime.thread_data.device_name}')
+        if not task.lock.acquire(blocking=False): raise Exception('Got locked task from queue.')
        try:
-            task.lock.acquire(blocking=False)
+            if runtime.thread_data.device == 'cpu' and is_alive() > 1:
+                # CPU is not the only device. Keep track of active time to unload resources later.
+                runtime.thread_data.lastActive = time.time()
+            # Open data generator.
            res = runtime.mk_img(task.request)
            if current_model_path == task.request.use_stable_diffusion_model:
                current_state = ServerStates.Rendering
            else:
                current_state = ServerStates.LoadingModel
+            # Start reading from generator.
+            dataQueue = None
+            if task.request.stream_progress_updates:
+                dataQueue = task.buffer_queue
+            for result in res:
+                if current_state == ServerStates.LoadingModel:
+                    current_state = ServerStates.Rendering
+                    current_model_path = task.request.use_stable_diffusion_model
+                    current_vae_path = task.request.use_vae_model
+                if isinstance(current_state_error, SystemExit) or isinstance(current_state_error, StopAsyncIteration) or isinstance(task.error, StopAsyncIteration):
+                    runtime.thread_data.stop_processing = True
+                    if isinstance(current_state_error, StopAsyncIteration):
+                        task.error = current_state_error
+                        current_state_error = None
+                        print(f'Session {task.request.session_id} sent cancel signal for task {id(task)}')
+                if dataQueue:
+                    dataQueue.put(result)
+                if isinstance(result, str):
+                    result = json.loads(result)
+                task.response = result
+                if 'output' in result:
+                    for out_obj in result['output']:
+                        if 'path' in out_obj:
+                            img_id = out_obj['path'][out_obj['path'].rindex('/') + 1:]
+                            task.temp_images[int(img_id)] = runtime.thread_data.temp_images[out_obj['path'][11:]]
+                        elif 'data' in out_obj:
+                            buf = runtime.base64_str_to_buffer(out_obj['data'])
+                            task.temp_images[result['output'].index(out_obj)] = buf
+                # Before looping back to the generator, mark cache as still alive.
+                task_cache.keep(task.request.session_id, TASK_TTL)
        except Exception as e:
            task.error = e
-            task.lock.release()
-            tasks_queue.task_done()
            print(traceback.format_exc())
            continue
-        dataQueue = None
-        if task.request.stream_progress_updates:
-            dataQueue = task.buffer_queue
-        for result in res:
-            if current_state == ServerStates.LoadingModel:
-                current_state = ServerStates.Rendering
-                current_model_path = task.request.use_stable_diffusion_model
-            if isinstance(current_state_error, SystemExit) or isinstance(current_state_error, StopAsyncIteration) or isinstance(task.error, StopAsyncIteration):
-                runtime.stop_processing = True
-                if isinstance(current_state_error, StopAsyncIteration):
-                    task.error = current_state_error
-                    current_state_error = None
-                    print(f'Session {task.request.session_id} sent cancel signal for task {id(task)}')
-            if dataQueue:
-                dataQueue.put(result)
-            if isinstance(result, str):
-                result = json.loads(result)
-            task.response = result
-            if 'output' in result:
-                for out_obj in result['output']:
-                    if 'path' in out_obj:
-                        img_id = out_obj['path'][out_obj['path'].rindex('/') + 1:]
-                        task.temp_images[int(img_id)] = runtime.temp_images[out_obj['path'][11:]]
-                    elif 'data' in out_obj:
-                        task.temp_images[result['output'].index(out_obj)] = out_obj['data']
-            task_cache.keep(task.request.session_id, TASK_TTL)
-        # Task completed
-        task.lock.release()
-        tasks_queue.task_done()
+        finally:
+            # Task completed
+            task.lock.release()
        task_cache.keep(task.request.session_id, TASK_TTL)
        if isinstance(task.error, StopAsyncIteration):
            print(f'Session {task.request.session_id} task {id(task)} cancelled!')
        elif task.error is not None:
            print(f'Session {task.request.session_id} task {id(task)} failed!')
        else:
-            print(f'Session {task.request.session_id} task {id(task)} completed.')
+            print(f'Session {task.request.session_id} task {id(task)} completed by {runtime.thread_data.device_name}.')
        current_state = ServerStates.Online

-render_thread = threading.Thread(target=thread_render)
+def get_cached_task(session_id:str, update_ttl:bool=False):
+    # By calling keep before tryGet, wont discard if was expired.
+    if update_ttl and not task_cache.keep(session_id, TASK_TTL):
+        # Failed to keep task, already gone.
+        return None
+    return task_cache.tryGet(session_id)

-def start_render_thread():
-    # Start Rendering Thread
-    render_thread.daemon = True
-    render_thread.start()
+def get_devices():
+    devices = {
+        'all': {},
+        'active': {},
+    }
+
+    def get_device_info(device):
+        if device == 'cpu':
+            return {'name': device_manager.get_processor_name()}
+            
+        mem_free, mem_total = torch.cuda.mem_get_info(device)
+        mem_free /= float(10**9)
+        mem_total /= float(10**9)
+
+        return {
+            'name': torch.cuda.get_device_name(device),
+            'mem_free': mem_free,
+            'mem_total': mem_total,
+        }
+
+    # list the compatible devices
+    gpu_count = torch.cuda.device_count()
+    for device in range(gpu_count):
+        device = f'cuda:{device}'
+        if not device_manager.is_device_compatible(device):
+            continue
+
+        devices['all'].update({device: get_device_info(device)})
+
+    devices['all'].update({'cpu': get_device_info('cpu')})
+
+    # list the activated devices
+    if not manager_lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('get_devices' + ERR_LOCK_FAILED)
+    try:
+        for rthread in render_threads:
+            if not rthread.is_alive():
+                continue
+            weak_data = weak_thread_data.get(rthread)
+            if not weak_data or not 'device' in weak_data or not 'device_name' in weak_data:
+                continue
+            device = weak_data['device']
+            devices['active'].update({device: get_device_info(device)})
+    finally:
+        manager_lock.release()
+
+    return devices
+
+def is_alive(device=None):
+    if not manager_lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('is_alive' + ERR_LOCK_FAILED)
+    nbr_alive = 0
+    try:
+        for rthread in render_threads:
+            if device is not None:
+                weak_data = weak_thread_data.get(rthread)
+                if weak_data is None or not 'device' in weak_data or weak_data['device'] is None:
+                    continue
+                thread_device = weak_data['device']
+                if thread_device != device:
+                    continue
+            if rthread.is_alive():
+                nbr_alive += 1
+        return nbr_alive
+    finally:
+        manager_lock.release()
+
+def start_render_thread(device):
+    if not manager_lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('start_render_thread' + ERR_LOCK_FAILED)
+    print('Start new Rendering Thread on device', device)
+    try:
+        rthread = threading.Thread(target=thread_render, kwargs={'device': device})
+        rthread.daemon = True
+        rthread.name = THREAD_NAME_PREFIX + device
+        rthread.start()
+        render_threads.append(rthread)
+    finally:
+        manager_lock.release()
+    timeout = DEVICE_START_TIMEOUT
+    while not rthread.is_alive() or not rthread in weak_thread_data or not 'device' in weak_thread_data[rthread]:
+        if rthread in weak_thread_data and 'error' in weak_thread_data[rthread]:
+            print(rthread, device, 'error:', weak_thread_data[rthread]['error'])
+            return False
+        if timeout <= 0:
+            return False
+        timeout -= 1
+        time.sleep(1)
+    return True
+
+def stop_render_thread(device):
+    try:
+        device_manager.validate_device_id(device, log_prefix='stop_render_thread')
+    except:
+        print(traceback.format_exec())
+        return False
+
+    if not manager_lock.acquire(blocking=True, timeout=LOCK_TIMEOUT): raise Exception('stop_render_thread' + ERR_LOCK_FAILED)
+    print('Stopping Rendering Thread on device', device)
+
+    try:
+        thread_to_remove = None
+        for rthread in render_threads:
+            weak_data = weak_thread_data.get(rthread)
+            if weak_data is None or not 'device' in weak_data or weak_data['device'] is None:
+                continue
+            thread_device = weak_data['device']
+            if thread_device == device:
+                weak_data['alive'] = False
+                thread_to_remove = rthread
+                break
+        if thread_to_remove is not None:
+            render_threads.remove(rthread)
+            return True
+    finally:
+        manager_lock.release()
+
+    return False
+
+def update_render_threads(render_devices, active_devices):
+    devices_to_start, devices_to_stop = device_manager.get_device_delta(render_devices, active_devices)
+    print('devices_to_start', devices_to_start)
+    print('devices_to_stop', devices_to_stop)
+
+    for device in devices_to_stop:
+        if is_alive(device) <= 0:
+            print(device, 'is not alive')
+            continue
+        if not stop_render_thread(device):
+            print(device, 'could not stop render thread')
+
+    for device in devices_to_start:
+        if is_alive(device) >= 1:
+            print(device, 'already registered.')
+            continue
+        if not start_render_thread(device):
+            print(device, 'failed to start.')
+
+    if is_alive() <= 0: # No running devices, probably invalid user config.
+        raise EnvironmentError('ERROR: No active render devices! Please verify the "render_devices" value in config.json')
+
+    print('active devices', get_devices()['active'])

 def shutdown_event(): # Signal render thread to close on shutdown
    global current_state_error
    current_state_error = SystemExit('Application shutting down.')

 def render(req : ImageRequest):
-    if not render_thread.is_alive(): # Render thread is dead
+    if is_alive() <= 0: # Render thread is dead
        raise ChildProcessError('Rendering thread has died.')
    # Alive, check if task in cache
    task = task_cache.tryGet(req.session_id)
@ -277,12 +519,12 @@ def render(req : ImageRequest):
    r.sampler = req.sampler
    # r.allow_nsfw = req.allow_nsfw
    r.turbo = req.turbo
-    r.use_cpu = req.use_cpu
    r.use_full_precision = req.use_full_precision
    r.save_to_disk_path = req.save_to_disk_path
    r.use_upscale: str = req.use_upscale
    r.use_face_correction = req.use_face_correction
    r.use_stable_diffusion_model = req.use_stable_diffusion_model
+    r.use_vae_model = req.use_vae_model
    r.show_only_filtered_image = req.show_only_filtered_image
    r.output_format = req.output_format

@ -293,7 +535,14 @@ def render(req : ImageRequest):
        r.stream_image_progress = False

    new_task = RenderTask(r)
+
    if task_cache.put(r.session_id, new_task, TASK_TTL):
-        tasks_queue.put(new_task, block=True, timeout=30)
-        return new_task
+        # Use twice the normal timeout for adding user requests.
+        # Tries to force task_cache.put to fail before tasks_queue.put would. 
+        if manager_lock.acquire(blocking=True, timeout=LOCK_TIMEOUT * 2):
+            try:
+                tasks_queue.append(new_task)
+                return new_task
+            finally:
+                manager_lock.release()
    raise RuntimeError('Failed to add task to cache.')
--- a/ui/server.py
+++ b/ui/server.py
@ -1,3 +1,7 @@
+"""server.py: FastAPI SD-UI Web Host.
+Notes:
+    async endpoints always run on the main thread. Without they run on the thread pool.
+"""
 import json
 import traceback

@ -16,14 +20,27 @@ UI_PLUGINS_DIR = os.path.abspath(os.path.join(SD_DIR, '..', 'plugins', 'ui'))

 OUTPUT_DIRNAME = "Stable Diffusion UI" # in the user's home folder
 TASK_TTL = 15 * 60 # Discard last session's task timeout
+APP_CONFIG_DEFAULTS = {
+    # auto: selects the cuda device with the most free memory, cuda: use the currently active cuda device.
+    'render_devices': 'auto', # valid entries: 'auto', 'cpu' or 'cuda:N' (where N is a GPU index)
+    'update_branch': 'main',
+    'ui': {
+        'open_browser_on_start': True,
+    },
+}
+APP_CONFIG_DEFAULT_MODELS = [
+    # needed to support the legacy installations
+    'custom-model', # Check if user has a custom model, use it first.
+    'sd-v1-4', # Default fallback.
+]

 from fastapi import FastAPI, HTTPException
 from fastapi.staticfiles import StaticFiles
 from starlette.responses import FileResponse, JSONResponse, StreamingResponse
 from pydantic import BaseModel
 import logging
-import queue, threading, time
-from typing import Any, Generator, Hashable, Optional, Union
+#import queue, threading, time
+from typing import Any, Generator, Hashable, List, Optional, Union

 from sd_internal import Request, Response, task_manager

@ -42,171 +59,7 @@ NOCACHE_HEADERS={"Cache-Control": "no-cache, no-store, must-revalidate", "Pragma
 app.mount('/media', StaticFiles(directory=os.path.join(SD_UI_DIR, 'media')), name="media")
 app.mount('/plugins', StaticFiles(directory=UI_PLUGINS_DIR), name="plugins")

-class SetAppConfigRequest(BaseModel):
-    update_branch: str = "main"
-
-# needs to support the legacy installations
-def get_initial_model_to_load():
-    custom_weight_path = os.path.join(SD_DIR, 'custom-model.ckpt')
-    ckpt_to_use = "sd-v1-4" if not os.path.exists(custom_weight_path) else "custom-model"
-
-    ckpt_to_use = os.path.join(SD_DIR, ckpt_to_use)
-
-    config = getConfig()
-    if 'model' in config and 'stable-diffusion' in config['model']:
-        model_name = config['model']['stable-diffusion']
-        model_path = resolve_model_to_use(model_name)
-
-        if os.path.exists(model_path + '.ckpt'):
-            ckpt_to_use = model_path
-        else:
-            print('Could not find the configured custom model at:', model_path + '.ckpt', '. Using the default one:', ckpt_to_use + '.ckpt')
-    return ckpt_to_use
-
-def resolve_model_to_use(model_name):
-    if model_name in ('sd-v1-4', 'custom-model'):
-        model_path = os.path.join(MODELS_DIR, 'stable-diffusion', model_name)
-
-        legacy_model_path = os.path.join(SD_DIR, model_name)
-        if not os.path.exists(model_path + '.ckpt') and os.path.exists(legacy_model_path + '.ckpt'):
-            model_path = legacy_model_path
-    else:
-        model_path = os.path.join(MODELS_DIR, 'stable-diffusion', model_name)
-    return model_path
-
-@app.on_event("shutdown")
-def shutdown_event(): # Signal render thread to close on shutdown
-    task_manager.current_state_error = SystemExit('Application shutting down.')
-
-@app.get('/')
-def read_root():
-    return FileResponse(os.path.join(SD_UI_DIR, 'index.html'), headers=NOCACHE_HEADERS)
-
-@app.get('/ping') # Get server and optionally session status.
-def ping(session_id:str=None):
-    if not task_manager.render_thread.is_alive(): # Render thread is dead.
-        if task_manager.current_state_error: raise HTTPException(status_code=500, detail=str(task_manager.current_state_error))
-        raise HTTPException(status_code=500, detail='Render thread is dead.')
-    if task_manager.current_state_error and not isinstance(task_manager.current_state_error, StopAsyncIteration): raise HTTPException(status_code=500, detail=str(task_manager.current_state_error))
-    # Alive
-    response = {'status': str(task_manager.current_state)}
-    if session_id:
-        task = task_manager.task_cache.tryGet(session_id)
-        if task:
-            response['task'] = id(task)
-            if task.lock.locked():
-                response['session'] = 'running'
-            elif isinstance(task.error, StopAsyncIteration):
-                response['session'] = 'stopped'
-            elif task.error:
-                response['session'] = 'error'
-            elif not task.buffer_queue.empty():
-                response['session'] = 'buffer'
-            elif task.response:
-                response['session'] = 'completed'
-            else:
-                response['session'] = 'pending'
-    return JSONResponse(response, headers=NOCACHE_HEADERS)
-
-def save_model_to_config(model_name):
-    config = getConfig()
-    if 'model' not in config:
-        config['model'] = {}
-
-    config['model']['stable-diffusion'] = model_name
-    setConfig(config)
-
-@app.post('/render')
-def render(req : task_manager.ImageRequest):
-    try:
-        save_model_to_config(req.use_stable_diffusion_model)
-        req.use_stable_diffusion_model = resolve_model_to_use(req.use_stable_diffusion_model)
-        new_task = task_manager.render(req)
-        response = {
-            'status': str(task_manager.current_state), 
-            'queue': task_manager.tasks_queue.qsize(),
-            'stream': f'/image/stream/{req.session_id}/{id(new_task)}',
-            'task': id(new_task)
-        }
-        return JSONResponse(response, headers=NOCACHE_HEADERS)
-    except ChildProcessError as e: # Render thread is dead
-        raise HTTPException(status_code=500, detail=f'Rendering thread has died.') # HTTP500 Internal Server Error
-    except ConnectionRefusedError as e: # Unstarted task pending, deny queueing more than one.
-        raise HTTPException(status_code=503, detail=f'Session {req.session_id} has an already pending task.') # HTTP503 Service Unavailable
-    except Exception as e:
-        raise HTTPException(status_code=500, detail=str(e))
-
-@app.get('/image/stream/{session_id:str}/{task_id:int}')
-def stream(session_id:str, task_id:int):
-    #TODO Move to WebSockets ??
-    task = task_manager.task_cache.tryGet(session_id)
-    if not task: raise HTTPException(status_code=410, detail='No request received.') # HTTP410 Gone
-    if (id(task) != task_id): raise HTTPException(status_code=409, detail=f'Wrong task id received. Expected:{id(task)}, Received:{task_id}') # HTTP409 Conflict
-    if task.buffer_queue.empty() and not task.lock.locked():
-        if task.response:
-            #print(f'Session {session_id} sending cached response')
-            return JSONResponse(task.response, headers=NOCACHE_HEADERS)
-        raise HTTPException(status_code=425, detail='Too Early, task not started yet.') # HTTP425 Too Early
-    #print(f'Session {session_id} opened live render stream {id(task.buffer_queue)}')
-    return StreamingResponse(task.read_buffer_generator(), media_type='application/json')
-
-@app.get('/image/stop')
-def stop(session_id:str=None):
-    if not session_id:
-        if task_manager.current_state == task_manager.ServerStates.Online or task_manager.current_state == task_manager.ServerStates.Unavailable:
-            raise HTTPException(status_code=409, detail='Not currently running any tasks.') # HTTP409 Conflict
-        task_manager.current_state_error = StopAsyncIteration('')
-        return {'OK'}
-    task = task_manager.task_cache.tryGet(session_id)
-    if not task: raise HTTPException(status_code=404, detail=f'Session {session_id} has no active task.') # HTTP404 Not Found
-    if isinstance(task.error, StopAsyncIteration): raise HTTPException(status_code=409, detail=f'Session {session_id} task is already stopped.') # HTTP409 Conflict
-    task.error = StopAsyncIteration('')
-    return {'OK'}
-
-@app.get('/image/tmp/{session_id}/{img_id:int}')
-def get_image(session_id, img_id):
-    task = task_manager.task_cache.tryGet(session_id)
-    if not task: raise HTTPException(status_code=410, detail=f'Session {session_id} has not submitted a task.') # HTTP410 Gone
-    if not task.temp_images[img_id]: raise HTTPException(status_code=425, detail='Too Early, task data is not available yet.') # HTTP425 Too Early
-    try:
-        img_data = task.temp_images[img_id]
-        if isinstance(img_data, str):
-            return img_data
-        img_data.seek(0)
-        return StreamingResponse(img_data, media_type='image/jpeg')
-    except KeyError as e:
-        raise HTTPException(status_code=500, detail=str(e))
-
-@app.post('/app_config')
-async def setAppConfig(req : SetAppConfigRequest):
-    try:
-        config = {
-            'update_branch': req.update_branch
-        }
-
-        config_json_str = json.dumps(config)
-        config_bat_str = f'@set update_branch={req.update_branch}'
-        config_sh_str = f'export update_branch={req.update_branch}'
-
-        config_json_path = os.path.join(CONFIG_DIR, 'config.json')
-        config_bat_path = os.path.join(CONFIG_DIR, 'config.bat')
-        config_sh_path = os.path.join(CONFIG_DIR, 'config.sh')
-
-        with open(config_json_path, 'w', encoding='utf-8') as f:
-            f.write(config_json_str)
-
-        with open(config_bat_path, 'w', encoding='utf-8') as f:
-            f.write(config_bat_str)
-
-        with open(config_sh_path, 'w', encoding='utf-8') as f:
-            f.write(config_sh_str)
-
-        return {'OK'}
-    except Exception as e:
-        print(traceback.format_exc())
-        raise HTTPException(status_code=500, detail=str(e))
-
-def getConfig(default_val={}):
+def getConfig(default_val=APP_CONFIG_DEFAULTS):
    try:
        config_json_path = os.path.join(CONFIG_DIR, 'config.json')
        if not os.path.exists(config_json_path):
@ -219,41 +72,152 @@ def getConfig(default_val={}):
        return default_val

 def setConfig(config):
-    try:
+    try: # config.json
        config_json_path = os.path.join(CONFIG_DIR, 'config.json')
        with open(config_json_path, 'w', encoding='utf-8') as f:
-            return json.dump(config, f)
-    except Exception as e:
-        print(str(e))
+            json.dump(config, f)
+    except:
        print(traceback.format_exc())

+    try: # config.bat
+        config_bat_path = os.path.join(CONFIG_DIR, 'config.bat')
+        config_bat = []
+
+        if 'update_branch' in config:
+            config_bat.append(f"@set update_branch={config['update_branch']}")
+        if os.getenv('SD_UI_BIND_PORT') is not None:
+            config_bat.append(f"@set SD_UI_BIND_PORT={os.getenv('SD_UI_BIND_PORT')}")
+        if os.getenv('SD_UI_BIND_IP') is not None:
+            config_bat.append(f"@set SD_UI_BIND_IP={os.getenv('SD_UI_BIND_IP')}")
+
+        if len(config_bat) > 0:
+            with open(config_bat_path, 'w', encoding='utf-8') as f:
+                f.write('\r\n'.join(config_bat))
+    except:
+        print(traceback.format_exc())
+
+    try: # config.sh
+        config_sh_path = os.path.join(CONFIG_DIR, 'config.sh')
+        config_sh = ['#!/bin/bash']
+
+        if 'update_branch' in config:
+            config_sh.append(f"export update_branch={config['update_branch']}")
+        if os.getenv('SD_UI_BIND_PORT') is not None:
+            config_sh.append(f"export SD_UI_BIND_PORT={os.getenv('SD_UI_BIND_PORT')}")
+        if os.getenv('SD_UI_BIND_IP') is not None:
+            config_sh.append(f"export SD_UI_BIND_IP={os.getenv('SD_UI_BIND_IP')}")
+
+        if len(config_sh) > 1:
+            with open(config_sh_path, 'w', encoding='utf-8') as f:
+                f.write('\n'.join(config_sh))
+    except:
+        print(traceback.format_exc())
+
+def resolve_model_to_use(model_name:str, model_type:str, model_dir:str, model_extensions:list, default_models=[]):
+    model_dirs = [os.path.join(MODELS_DIR, model_dir), SD_DIR]
+    if not model_name: # When None try user configured model.
+        config = getConfig()
+        if 'model' in config and model_type in config['model']:
+            model_name = config['model'][model_type]
+    if model_name:
+        # Check models directory
+        models_dir_path = os.path.join(MODELS_DIR, model_dir, model_name)
+        for model_extension in model_extensions:
+            if os.path.exists(models_dir_path + model_extension):
+                return models_dir_path
+            if os.path.exists(model_name + model_extension):
+                # Direct Path to file
+                model_name = os.path.abspath(model_name)
+                return model_name
+    # Default locations
+    if model_name in default_models:
+        default_model_path = os.path.join(SD_DIR, model_name)
+        for model_extension in model_extensions:
+            if os.path.exists(default_model_path + model_extension):
+                return default_model_path
+    # Can't find requested model, check the default paths.
+    for default_model in default_models:
+        for model_dir in model_dirs:
+            default_model_path = os.path.join(model_dir, default_model)
+            for model_extension in model_extensions:
+                if os.path.exists(default_model_path + model_extension):
+                    if model_name is not None:
+                        print(f'Could not find the configured custom model {model_name}{model_extension}. Using the default one: {default_model_path}{model_extension}')
+                    return default_model_path
+    raise Exception('No valid models found.')
+
+def resolve_ckpt_to_use(model_name:str=None):
+    return resolve_model_to_use(model_name, model_type='stable-diffusion', model_dir='stable-diffusion', model_extensions=['.ckpt'], default_models=APP_CONFIG_DEFAULT_MODELS)
+
+def resolve_vae_to_use(model_name:str=None):
+    try:
+        return resolve_model_to_use(model_name, model_type='vae', model_dir='vae', model_extensions=['.vae.pt', '.ckpt'], default_models=[])
+    except:
+        return None
+
+class SetAppConfigRequest(BaseModel):
+    update_branch: str = None
+    render_devices: Union[List[str], List[int], str, int] = None
+    model_vae: str = None
+    ui_open_browser_on_start: bool = None
+
+@app.post('/app_config')
+async def setAppConfig(req : SetAppConfigRequest):
+    config = getConfig()
+    if req.update_branch is not None:
+        config['update_branch'] = req.update_branch
+    if req.render_devices is not None:
+        update_render_devices_in_config(config, req.render_devices)
+    if req.ui_open_browser_on_start is not None:
+        if 'ui' not in config:
+            config['ui'] = {}
+        config['ui']['open_browser_on_start'] = req.ui_open_browser_on_start
+    try:
+        setConfig(config)
+
+        if req.render_devices:
+            update_render_threads()
+
+        return JSONResponse({'status': 'OK'}, headers=NOCACHE_HEADERS)
+    except Exception as e:
+        print(traceback.format_exc())
+        raise HTTPException(status_code=500, detail=str(e))
+
 def getModels():
    models = {
        'active': {
            'stable-diffusion': 'sd-v1-4',
+            'vae': '',
        },
        'options': {
            'stable-diffusion': ['sd-v1-4'],
+            'vae': [],
        },
    }

+    def listModels(models_dirname, model_type, model_extensions):
+        models_dir = os.path.join(MODELS_DIR, models_dirname)
+        if not os.path.exists(models_dir):
+            os.makedirs(models_dir)
+
+        for file in os.listdir(models_dir):
+            for model_extension in model_extensions:
+                if file.endswith(model_extension):
+                    model_name = file[:-len(model_extension)]
+                    models['options'][model_type].append(model_name)
+
+        models['options'][model_type] = [*set(models['options'][model_type])] # remove duplicates
+        models['options'][model_type].sort()
+
    # custom models
-    sd_models_dir = os.path.join(MODELS_DIR, 'stable-diffusion')
-    for file in os.listdir(sd_models_dir):
-        if file.endswith('.ckpt'):
-            model_name = os.path.splitext(file)[0]
-            models['options']['stable-diffusion'].append(model_name)
+    listModels(models_dirname='stable-diffusion', model_type='stable-diffusion', model_extensions=['.ckpt'])
+    listModels(models_dirname='vae', model_type='vae', model_extensions=['.vae.pt', '.ckpt'])

    # legacy
    custom_weight_path = os.path.join(SD_DIR, 'custom-model.ckpt')
    if os.path.exists(custom_weight_path):
-        models['active']['stable-diffusion'] = 'custom-model'
        models['options']['stable-diffusion'].append('custom-model')

-    config = getConfig()
-    if 'model' in config and 'stable-diffusion' in config['model']:
-        models['active']['stable-diffusion'] = config['model']['stable-diffusion']
-
    return models

 def getUIPlugins():
@ -272,8 +236,13 @@ def read_web_data(key:str=None):
    elif key == 'app_config':
        config = getConfig(default_val=None)
        if config is None:
-            raise HTTPException(status_code=500, detail="Config file is missing or unreadable")
+            config = APP_CONFIG_DEFAULTS
        return JSONResponse(config, headers=NOCACHE_HEADERS)
+    elif key == 'devices':
+        config = getConfig()
+        devices = task_manager.get_devices()
+        devices['config'] = config.get('render_devices', "auto")
+        return JSONResponse(devices, headers=NOCACHE_HEADERS)
    elif key == 'models':
        return JSONResponse(getModels(), headers=NOCACHE_HEADERS)
    elif key == 'modifiers': return FileResponse(os.path.join(SD_UI_DIR, 'modifiers.json'), headers=NOCACHE_HEADERS)
@ -282,6 +251,123 @@ def read_web_data(key:str=None):
    else:
        raise HTTPException(status_code=404, detail=f'Request for unknown {key}') # HTTP404 Not Found

+@app.get('/ping') # Get server and optionally session status.
+def ping(session_id:str=None):
+    if task_manager.is_alive() <= 0: # Check that render threads are alive.
+        if task_manager.current_state_error: raise HTTPException(status_code=500, detail=str(task_manager.current_state_error))
+        raise HTTPException(status_code=500, detail='Render thread is dead.')
+    if task_manager.current_state_error and not isinstance(task_manager.current_state_error, StopAsyncIteration): raise HTTPException(status_code=500, detail=str(task_manager.current_state_error))
+    # Alive
+    response = {'status': str(task_manager.current_state)}
+    if session_id:
+        task = task_manager.get_cached_task(session_id, update_ttl=True)
+        if task:
+            response['task'] = id(task)
+            if task.lock.locked():
+                response['session'] = 'running'
+            elif isinstance(task.error, StopAsyncIteration):
+                response['session'] = 'stopped'
+            elif task.error:
+                response['session'] = 'error'
+            elif not task.buffer_queue.empty():
+                response['session'] = 'buffer'
+            elif task.response:
+                response['session'] = 'completed'
+            else:
+                response['session'] = 'pending'
+    response['devices'] = task_manager.get_devices()
+    return JSONResponse(response, headers=NOCACHE_HEADERS)
+
+def save_model_to_config(ckpt_model_name, vae_model_name):
+    config = getConfig()
+    if 'model' not in config:
+        config['model'] = {}
+
+    config['model']['stable-diffusion'] = ckpt_model_name
+    config['model']['vae'] = vae_model_name
+
+    if vae_model_name is None or vae_model_name == "":
+        del config['model']['vae']
+
+    setConfig(config)
+
+def update_render_devices_in_config(config, render_devices):
+    if render_devices not in ('cpu', 'auto') and not render_devices.startswith('cuda:'):
+        raise HTTPException(status_code=400, detail=f'Invalid render device requested: {render_devices}')
+
+    if render_devices.startswith('cuda:'):
+        render_devices = render_devices.split(',')
+
+    config['render_devices'] = render_devices
+
+@app.post('/render')
+def render(req : task_manager.ImageRequest):
+    try:
+        save_model_to_config(req.use_stable_diffusion_model, req.use_vae_model)
+        req.use_stable_diffusion_model = resolve_ckpt_to_use(req.use_stable_diffusion_model)
+        req.use_vae_model = resolve_vae_to_use(req.use_vae_model)
+        new_task = task_manager.render(req)
+        response = {
+            'status': str(task_manager.current_state), 
+            'queue': len(task_manager.tasks_queue),
+            'stream': f'/image/stream/{req.session_id}/{id(new_task)}',
+            'task': id(new_task)
+        }
+        return JSONResponse(response, headers=NOCACHE_HEADERS)
+    except ChildProcessError as e: # Render thread is dead
+        raise HTTPException(status_code=500, detail=f'Rendering thread has died.') # HTTP500 Internal Server Error
+    except ConnectionRefusedError as e: # Unstarted task pending, deny queueing more than one.
+        raise HTTPException(status_code=503, detail=f'Session {req.session_id} has an already pending task.') # HTTP503 Service Unavailable
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+@app.get('/image/stream/{session_id:str}/{task_id:int}')
+def stream(session_id:str, task_id:int):
+    #TODO Move to WebSockets ??
+    task = task_manager.get_cached_task(session_id, update_ttl=True)
+    if not task: raise HTTPException(status_code=410, detail='No request received.') # HTTP410 Gone
+    if (id(task) != task_id): raise HTTPException(status_code=409, detail=f'Wrong task id received. Expected:{id(task)}, Received:{task_id}') # HTTP409 Conflict
+    if task.buffer_queue.empty() and not task.lock.locked():
+        if task.response:
+            #print(f'Session {session_id} sending cached response')
+            return JSONResponse(task.response, headers=NOCACHE_HEADERS)
+        raise HTTPException(status_code=425, detail='Too Early, task not started yet.') # HTTP425 Too Early
+    #print(f'Session {session_id} opened live render stream {id(task.buffer_queue)}')
+    return StreamingResponse(task.read_buffer_generator(), media_type='application/json')
+
+@app.get('/image/stop')
+def stop(session_id:str=None):
+    if not session_id:
+        if task_manager.current_state == task_manager.ServerStates.Online or task_manager.current_state == task_manager.ServerStates.Unavailable:
+            raise HTTPException(status_code=409, detail='Not currently running any tasks.') # HTTP409 Conflict
+        task_manager.current_state_error = StopAsyncIteration('')
+        return {'OK'}
+    task = task_manager.get_cached_task(session_id, update_ttl=False)
+    if not task: raise HTTPException(status_code=404, detail=f'Session {session_id} has no active task.') # HTTP404 Not Found
+    if isinstance(task.error, StopAsyncIteration): raise HTTPException(status_code=409, detail=f'Session {session_id} task is already stopped.') # HTTP409 Conflict
+    task.error = StopAsyncIteration('')
+    return {'OK'}
+
+@app.get('/image/tmp/{session_id}/{img_id:int}')
+def get_image(session_id, img_id):
+    task = task_manager.get_cached_task(session_id, update_ttl=True)
+    if not task: raise HTTPException(status_code=410, detail=f'Session {session_id} has not submitted a task.') # HTTP410 Gone
+    if not task.temp_images[img_id]: raise HTTPException(status_code=425, detail='Too Early, task data is not available yet.') # HTTP425 Too Early
+    try:
+        img_data = task.temp_images[img_id]
+        img_data.seek(0)
+        return StreamingResponse(img_data, media_type='image/jpeg')
+    except KeyError as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+@app.get('/')
+def read_root():
+    return FileResponse(os.path.join(SD_UI_DIR, 'index.html'), headers=NOCACHE_HEADERS)
+
+@app.on_event("shutdown")
+def shutdown_event(): # Signal render thread to close on shutdown
+    task_manager.current_state_error = SystemExit('Application shutting down.')
+
 # don't log certain requests
 class LogSuppressFilter(logging.Filter):
    def filter(self, record: logging.LogRecord) -> bool:
@ -292,8 +378,25 @@ class LogSuppressFilter(logging.Filter):
        return True
 logging.getLogger('uvicorn.access').addFilter(LogSuppressFilter())

-task_manager.default_model_to_load = get_initial_model_to_load()
-task_manager.start_render_thread()
+# Start the task_manager
+task_manager.default_model_to_load = resolve_ckpt_to_use()
+task_manager.default_vae_to_load = resolve_vae_to_use()
+
+def update_render_threads():
+    config = getConfig()
+    render_devices = config.get('render_devices', 'auto')
+    active_devices = task_manager.get_devices()['active'].keys()
+
+    print('requesting for render_devices', render_devices)
+    task_manager.update_render_threads(render_devices, active_devices)
+
+update_render_threads()

 # start the browser ui
-import webbrowser; webbrowser.open('http://localhost:9000')
+def open_browser():
+    config = getConfig()
+    ui = config.get('ui', {})
+    if ui.get('open_browser_on_start', True):
+        import webbrowser; webbrowser.open('http://localhost:9000')
+
+open_browser()