Merge branch 'main' into remove-deprecation-raise

.changeset/better-parents-bow.md

@@ -0,0 +1,6 @@
+---
+"@gradio/app": patch
+"gradio": patch
+---
+
+fix:Fix updating interactive prop

.changeset/cute-poems-cover.md

@@ -0,0 +1,5 @@
+---
+"gradio": minor
+---
+
+feat:Remove session if browser closed on mobile

.changeset/fifty-baboons-shout.md

@@ -0,0 +1,6 @@
+---
+"@gradio/radio": patch
+"gradio": patch
+---
+
+fix:Fix bug where radio.select passes the previous value to the function instead of the selected value

.changeset/fruity-peaches-appear.md

@@ -0,0 +1,6 @@
+---
+"@gradio/model3d": minor
+"gradio": minor
+---
+
+feat:Model3D panning, improved UX

.changeset/hot-donuts-notice.md

@@ -0,0 +1,5 @@
+---
+"website": patch
+---
+
+feat:Some tweaks to the Custom Components Guide

.changeset/kind-toys-stop.md

@@ -0,0 +1,5 @@
+---
+"gradio": minor
+---
+
+feat:Fixes input `Image` component with `streaming=True`

.changeset/major-bushes-ring.md

@@ -0,0 +1,5 @@
+---
+"gradio": patch
+---
+
+fix:Add likeable to config for Chatbot

.changeset/polite-radios-kiss.md

@@ -0,0 +1,10 @@
+---
+"@gradio/app": patch
+"@gradio/audio": patch
+"@gradio/tootils": patch
+"@gradio/upload": patch
+"@gradio/video": patch
+"gradio": patch
+---
+
+fix:Video/Audio fixes

.changeset/pretty-doodles-work.md

@@ -0,0 +1,7 @@
+---
+"@gradio/radio": patch
+"@gradio/tootils": patch
+"gradio": patch
+---
+
+fix:ensure radios have different names

.changeset/public-banks-film.md

@@ -0,0 +1,5 @@
+---
+"gradio": patch
+---
+
+fix:Fixes: Initial message is overwrtitten in chat interface 

.changeset/public-hotels-divide.md

@@ -0,0 +1,8 @@
+---
+"@gradio/app": patch
+"@gradio/checkboxgroup": patch
+"@gradio/tootils": patch
+"gradio": patch
+---
+
+fix:Ensure `gr.CheckboxGroup` updates as expected.

.changeset/purple-areas-see.md

@@ -0,0 +1,7 @@
+---
+"@gradio/app": minor
+"@gradio/preview": minor
+"gradio": minor
+---
+
+feat:Improve Embed and CDN handling and fix a couple of related bugs 

.changeset/quick-cases-stay.md

@@ -0,0 +1,6 @@
+---
+"gradio": minor
+"website": minor
+---
+
+feat:Fix various issues with demos on website

.changeset/sour-sheep-look.md

@@ -0,0 +1,6 @@
+---
+"@gradio/model3d": minor
+"gradio": minor
+---
+
+feat:Ensure Model 3D updates when attributes change

.changeset/upset-radios-switch.md

@@ -0,0 +1,6 @@
+---
+"@gradio/fileexplorer": patch
+"gradio": patch
+---
+
+fix:Fix file overflow and add keyboard navigation to `FileExplorer`

.config/.prettierignore

@@ -24,4 +24,7 @@
 sweep.yaml
 **/.vercel/**
 **/build/**
-**/*.md
+**/*.md
+**/src/lib/json/**/*
+**/playwright/.cache/**/*
+**/theme/src/pollen.css

.config/basevite.config.ts

@@ -27,10 +27,7 @@ const version = JSON.parse(readFileSync(version_path, { encoding: "utf-8" }))
 
 //@ts-ignore
 export default defineConfig(({ mode }) => {
-	const production =
-		mode === "production:cdn" ||
-		mode === "production:local" ||
-		mode === "production:website";
+	const production = mode === "production";
 
 	return {
 		server: {

.github/workflows/deploy-pr-to-spaces.yml

@@ -62,7 +62,7 @@ jobs:
         github.event.workflow_run.conclusion == 'success'
       id: upload-website-demos
       run: |
-        python scripts/upload_website_demos.py --AUTH_TOKEN ${{ secrets.SPACES_DEPLOY_TOKEN }} \
+        python scripts/upload_website_demos.py --AUTH_TOKEN ${{ secrets.WEBSITE_SPACES_DEPLOY_TOKEN }} \
         --WHEEL_URL https://gradio-builds.s3.amazonaws.com/${{ steps.set-outputs.outputs.gh_sha }}/ \
         --GRADIO_VERSION ${{ steps.set-outputs.outputs.gradio_version }}
 

.github/workflows/ui.yml

@@ -30,6 +30,8 @@ jobs:
         run: pnpm --filter @gradio/client build
       - name: build the wasm module
         run: pnpm --filter @gradio/wasm build
+      - name: format check
+        run: pnpm format:check
       - name: lint
         run: pnpm lint
       - name: typecheck

.vscode/settings.json

@@ -18,5 +18,6 @@
 	"eslint.experimental.useFlatConfig": true,
 	"eslint.options": {
 		"overrideConfigFile": "./.config/eslint.config.js"
-	}
+	},
+	"typescript.tsdk": "node_modules/typescript/lib"
 }

README.md

@@ -163,7 +163,7 @@ def sepia(input_img):
     sepia_img /= sepia_img.max()
     return sepia_img
 
-demo = gr.Interface(sepia, gr.Image(shape=(200, 200)), "image")
+demo = gr.Interface(sepia, gr.Image(width=200, height=200), "image")
 demo.launch()
 ```
 

build_pypi.sh

@@ -9,12 +9,9 @@ new_version=$(python -c "import json; f = open('$FILE', 'r'); data = json.load(f
 GRADIO_VERSION=$new_version
 
 rm -rf gradio/templates/frontend
-rm -rf gradio/templates/cdn
 pnpm i --frozen-lockfile --ignore-scripts
 GRADIO_VERSION=$new_version pnpm build
-GRADIO_VERSION=$new_version pnpm build:cdn
-aws s3 cp gradio/templates/cdn "s3://gradio/${new_version}/" --recursive --region us-west-2
-cp gradio/templates/cdn/index.html gradio/templates/frontend/share.html
+aws s3 cp gradio/templates/frontend "s3://gradio/${new_version}/" --recursive --region us-west-2
 
 rm -rf dist/*
 rm -rf build/*

demo/Echocardiogram-Segmentation/run.ipynb

@@ -1 +1 @@
-{"cells": [{"cell_type": "markdown", "id": "302934307671667531413257853548643485645", "metadata": {}, "source": ["# Gradio Demo: Echocardiogram-Segmentation"]}, {"cell_type": "code", "execution_count": null, "id": "272996653310673477252411125948039410165", "metadata": {}, "outputs": [], "source": ["!pip install -q gradio -f https://download.pytorch.org/whl/torch_stable.html numpy matplotlib wget torch torchvision   "]}, {"cell_type": "code", "execution_count": null, "id": "288918539441861185822528903084949547379", "metadata": {}, "outputs": [], "source": ["# Downloading files from the demo repo\n", "import os\n", "!wget -q https://github.com/gradio-app/gradio/raw/main/demo/Echocardiogram-Segmentation/img1.jpg\n", "!wget -q https://github.com/gradio-app/gradio/raw/main/demo/Echocardiogram-Segmentation/img2.jpg"]}, {"cell_type": "code", "execution_count": null, "id": "44380577570523278879349135829904343037", "metadata": {}, "outputs": [], "source": ["import os\n", "import numpy as np\n", "import torch\n", "import torchvision\n", "import wget \n", "\n", "\n", "destination_folder = \"output\"\n", "destination_for_weights = \"weights\"\n", "\n", "if os.path.exists(destination_for_weights):\n", "    print(\"The weights are at\", destination_for_weights)\n", "else:\n", "    print(\"Creating folder at \", destination_for_weights, \" to store weights\")\n", "    os.mkdir(destination_for_weights)\n", "    \n", "segmentationWeightsURL = 'https://github.com/douyang/EchoNetDynamic/releases/download/v1.0.0/deeplabv3_resnet50_random.pt'\n", "\n", "if not os.path.exists(os.path.join(destination_for_weights, os.path.basename(segmentationWeightsURL))):\n", "    print(\"Downloading Segmentation Weights, \", segmentationWeightsURL,\" to \",os.path.join(destination_for_weights, os.path.basename(segmentationWeightsURL)))\n", "    filename = wget.download(segmentationWeightsURL, out = destination_for_weights)\n", "else:\n", "    print(\"Segmentation Weights already present\")\n", "\n", "torch.cuda.empty_cache()\n", "\n", "def collate_fn(x):\n", "    x, f = zip(*x)\n", "    i = list(map(lambda t: t.shape[1], x))\n", "    x = torch.as_tensor(np.swapaxes(np.concatenate(x, 1), 0, 1))\n", "    return x, f, i\n", "\n", "model = torchvision.models.segmentation.deeplabv3_resnet50(pretrained=False, aux_loss=False)\n", "model.classifier[-1] = torch.nn.Conv2d(model.classifier[-1].in_channels, 1, kernel_size=model.classifier[-1].kernel_size)\n", "\n", "print(\"loading weights from \", os.path.join(destination_for_weights, \"deeplabv3_resnet50_random\"))\n", "\n", "if torch.cuda.is_available():\n", "    print(\"cuda is available, original weights\")\n", "    device = torch.device(\"cuda\")\n", "    model = torch.nn.DataParallel(model)\n", "    model.to(device)\n", "    checkpoint = torch.load(os.path.join(destination_for_weights, os.path.basename(segmentationWeightsURL)))\n", "    model.load_state_dict(checkpoint['state_dict'])\n", "else:\n", "    print(\"cuda is not available, cpu weights\")\n", "    device = torch.device(\"cpu\")\n", "    checkpoint = torch.load(os.path.join(destination_for_weights, os.path.basename(segmentationWeightsURL)), map_location = \"cpu\")\n", "    state_dict_cpu = {k[7:]: v for (k, v) in checkpoint['state_dict'].items()}\n", "    model.load_state_dict(state_dict_cpu)\n", "\n", "model.eval()\n", "\n", "def segment(input):\n", "    inp = input\n", "    x = inp.transpose([2, 0, 1])  #  channels-first\n", "    x = np.expand_dims(x, axis=0)  # adding a batch dimension    \n", "    \n", "    mean = x.mean(axis=(0, 2, 3))\n", "    std = x.std(axis=(0, 2, 3))\n", "    x = x - mean.reshape(1, 3, 1, 1)\n", "    x = x / std.reshape(1, 3, 1, 1)\n", "    \n", "    with torch.no_grad():\n", "        x = torch.from_numpy(x).type('torch.FloatTensor').to(device)\n", "        output = model(x)    \n", "    \n", "    y = output['out'].numpy()\n", "    y = y.squeeze()\n", "    \n", "    out = y>0    \n", "    \n", "    mask = inp.copy()\n", "    mask[out] = np.array([0, 0, 255])\n", "    \n", "    return mask\n", "\n", "import gradio as gr\n", "\n", "i = gr.Image(shape=(112, 112), label=\"Echocardiogram\")\n", "o = gr.Image(label=\"Segmentation Mask\")\n", "\n", "examples = [[\"img1.jpg\"], [\"img2.jpg\"]]\n", "title = None #\"Left Ventricle Segmentation\"\n", "description = \"This semantic segmentation model identifies the left ventricle in echocardiogram images.\"\n", "# videos. Accurate evaluation of the motion and size of the left ventricle is crucial for the assessment of cardiac function and ejection fraction. In this interface, the user inputs apical-4-chamber images from echocardiography videos and the model will output a prediction of the localization of the left ventricle in blue. This model was trained on the publicly released EchoNet-Dynamic dataset of 10k echocardiogram videos with 20k expert annotations of the left ventricle and published as part of \u2018Video-based AI for beat-to-beat assessment of cardiac function\u2019 by Ouyang et al. in Nature, 2020.\"\n", "thumbnail = \"https://raw.githubusercontent.com/gradio-app/hub-echonet/master/thumbnail.png\"\n", "gr.Interface(segment, i, o, examples=examples, allow_flagging=False, analytics_enabled=False, thumbnail=thumbnail, cache_examples=False).launch()\n"]}], "metadata": {}, "nbformat": 4, "nbformat_minor": 5}
+{"cells": [{"cell_type": "markdown", "id": "302934307671667531413257853548643485645", "metadata": {}, "source": ["# Gradio Demo: Echocardiogram-Segmentation"]}, {"cell_type": "code", "execution_count": null, "id": "272996653310673477252411125948039410165", "metadata": {}, "outputs": [], "source": ["!pip install -q gradio -f https://download.pytorch.org/whl/torch_stable.html numpy matplotlib wget torch torchvision   "]}, {"cell_type": "code", "execution_count": null, "id": "288918539441861185822528903084949547379", "metadata": {}, "outputs": [], "source": ["# Downloading files from the demo repo\n", "import os\n", "!wget -q https://github.com/gradio-app/gradio/raw/main/demo/Echocardiogram-Segmentation/img1.jpg\n", "!wget -q https://github.com/gradio-app/gradio/raw/main/demo/Echocardiogram-Segmentation/img2.jpg"]}, {"cell_type": "code", "execution_count": null, "id": "44380577570523278879349135829904343037", "metadata": {}, "outputs": [], "source": ["import os\n", "import numpy as np\n", "import torch\n", "import torchvision\n", "import wget \n", "\n", "\n", "destination_folder = \"output\"\n", "destination_for_weights = \"weights\"\n", "\n", "if os.path.exists(destination_for_weights):\n", "    print(\"The weights are at\", destination_for_weights)\n", "else:\n", "    print(\"Creating folder at \", destination_for_weights, \" to store weights\")\n", "    os.mkdir(destination_for_weights)\n", "    \n", "segmentationWeightsURL = 'https://github.com/douyang/EchoNetDynamic/releases/download/v1.0.0/deeplabv3_resnet50_random.pt'\n", "\n", "if not os.path.exists(os.path.join(destination_for_weights, os.path.basename(segmentationWeightsURL))):\n", "    print(\"Downloading Segmentation Weights, \", segmentationWeightsURL,\" to \",os.path.join(destination_for_weights, os.path.basename(segmentationWeightsURL)))\n", "    filename = wget.download(segmentationWeightsURL, out = destination_for_weights)\n", "else:\n", "    print(\"Segmentation Weights already present\")\n", "\n", "torch.cuda.empty_cache()\n", "\n", "def collate_fn(x):\n", "    x, f = zip(*x)\n", "    i = list(map(lambda t: t.shape[1], x))\n", "    x = torch.as_tensor(np.swapaxes(np.concatenate(x, 1), 0, 1))\n", "    return x, f, i\n", "\n", "model = torchvision.models.segmentation.deeplabv3_resnet50(pretrained=False, aux_loss=False)\n", "model.classifier[-1] = torch.nn.Conv2d(model.classifier[-1].in_channels, 1, kernel_size=model.classifier[-1].kernel_size)\n", "\n", "print(\"loading weights from \", os.path.join(destination_for_weights, \"deeplabv3_resnet50_random\"))\n", "\n", "if torch.cuda.is_available():\n", "    print(\"cuda is available, original weights\")\n", "    device = torch.device(\"cuda\")\n", "    model = torch.nn.DataParallel(model)\n", "    model.to(device)\n", "    checkpoint = torch.load(os.path.join(destination_for_weights, os.path.basename(segmentationWeightsURL)))\n", "    model.load_state_dict(checkpoint['state_dict'])\n", "else:\n", "    print(\"cuda is not available, cpu weights\")\n", "    device = torch.device(\"cpu\")\n", "    checkpoint = torch.load(os.path.join(destination_for_weights, os.path.basename(segmentationWeightsURL)), map_location = \"cpu\")\n", "    state_dict_cpu = {k[7:]: v for (k, v) in checkpoint['state_dict'].items()}\n", "    model.load_state_dict(state_dict_cpu)\n", "\n", "model.eval()\n", "\n", "def segment(input):\n", "    inp = input\n", "    x = inp.transpose([2, 0, 1])  #  channels-first\n", "    x = np.expand_dims(x, axis=0)  # adding a batch dimension    \n", "    \n", "    mean = x.mean(axis=(0, 2, 3))\n", "    std = x.std(axis=(0, 2, 3))\n", "    x = x - mean.reshape(1, 3, 1, 1)\n", "    x = x / std.reshape(1, 3, 1, 1)\n", "    \n", "    with torch.no_grad():\n", "        x = torch.from_numpy(x).type('torch.FloatTensor').to(device)\n", "        output = model(x)    \n", "    \n", "    y = output['out'].numpy()\n", "    y = y.squeeze()\n", "    \n", "    out = y>0    \n", "    \n", "    mask = inp.copy()\n", "    mask[out] = np.array([0, 0, 255])\n", "    \n", "    return mask\n", "\n", "import gradio as gr\n", "\n", "i = gr.Image(label=\"Echocardiogram\")\n", "o = gr.Image(label=\"Segmentation Mask\")\n", "\n", "examples = [[\"img1.jpg\"], [\"img2.jpg\"]]\n", "title = None #\"Left Ventricle Segmentation\"\n", "description = \"This semantic segmentation model identifies the left ventricle in echocardiogram images.\"\n", "# videos. Accurate evaluation of the motion and size of the left ventricle is crucial for the assessment of cardiac function and ejection fraction. In this interface, the user inputs apical-4-chamber images from echocardiography videos and the model will output a prediction of the localization of the left ventricle in blue. This model was trained on the publicly released EchoNet-Dynamic dataset of 10k echocardiogram videos with 20k expert annotations of the left ventricle and published as part of \u2018Video-based AI for beat-to-beat assessment of cardiac function\u2019 by Ouyang et al. in Nature, 2020.\"\n", "thumbnail = \"https://raw.githubusercontent.com/gradio-app/hub-echonet/master/thumbnail.png\"\n", "gr.Interface(segment, i, o, examples=examples, analytics_enabled=False, thumbnail=thumbnail, cache_examples=False).launch()\n"]}], "metadata": {}, "nbformat": 4, "nbformat_minor": 5}

demo/Echocardiogram-Segmentation/run.py

@@ -77,12 +77,12 @@ def segment(input):
 
 import gradio as gr
 
-i = gr.Image(shape=(112, 112), label="Echocardiogram")
+i = gr.Image(label="Echocardiogram")
 o = gr.Image(label="Segmentation Mask")
 
 examples = [["img1.jpg"], ["img2.jpg"]]
 title = None #"Left Ventricle Segmentation"
 description = "This semantic segmentation model identifies the left ventricle in echocardiogram images."
 # videos. Accurate evaluation of the motion and size of the left ventricle is crucial for the assessment of cardiac function and ejection fraction. In this interface, the user inputs apical-4-chamber images from echocardiography videos and the model will output a prediction of the localization of the left ventricle in blue. This model was trained on the publicly released EchoNet-Dynamic dataset of 10k echocardiogram videos with 20k expert annotations of the left ventricle and published as part of ‘Video-based AI for beat-to-beat assessment of cardiac function’ by Ouyang et al. in Nature, 2020."
 thumbnail = "https://raw.githubusercontent.com/gradio-app/hub-echonet/master/thumbnail.png"
-gr.Interface(segment, i, o, examples=examples, allow_flagging=False, analytics_enabled=False, thumbnail=thumbnail, cache_examples=False).launch()
+gr.Interface(segment, i, o, examples=examples, analytics_enabled=False, thumbnail=thumbnail, cache_examples=False).launch()