Add Gradio helpers - part 7 (openvinotoolkit#2334)

Ticket: CVS-147626

Notebooks:

1. photo-maker/photo-maker.ipynb
2. pix2struct-docvqa/pix2struct-docvqa.ipynb
3. pixart/pixart.ipynb
4. qrcode-monster/qrcode-monster.ipynb
5. riffusion-text-to-music/riffusion-text-to-music.ipynb
6. rmbg-background-removal/rmbg-background-removal.ipynb
7. sdxl-turbo/sdxl-turbo.ipynb
8. sketch-to-image-pix2pix-turbo/sketch-to-image-pix2pix-turbo.ipynb
9. softvc-voice-conversion/softvc-voice-conversion.ipynb
10. sound-generation-audioldm2/sound-generation-audioldm2.ipynb

Author: yatarkan

notebooks/photo-maker/gradio_helper.py

@@ -0,0 +1,61 @@
+from typing import Callable
+import gradio as gr
+
+
+trigger_word = "img"
+prompt = "sci-fi, closeup portrait photo of a man img in Iron man suit, face"
+negative_prompt = "(asymmetry, worst quality, low quality, illustration, 3d, 2d, painting, cartoons, sketch), open mouth"
+
+
+def make_demo(fn: Callable):
+    with gr.Blocks() as demo:
+        with gr.Column():
+            with gr.Row():
+                input_image = gr.Image(label="Your image", sources=["upload"], type="pil")
+                output_image = gr.Image(label="Generated Images", type="pil")
+            positive_input = gr.Textbox(label=f"Text prompt, Trigger words is '{trigger_word}'")
+            neg_input = gr.Textbox(label="Negative prompt")
+            with gr.Row():
+                seed_input = gr.Slider(0, 10_000_000, value=42, label="Seed")
+                steps_input = gr.Slider(label="Steps", value=10, minimum=5, maximum=50, step=1)
+                style_strength_ratio_input = gr.Slider(label="Style strength ratio", value=20, minimum=5, maximum=100, step=5)
+                btn = gr.Button()
+            btn.click(
+                fn=fn,
+                inputs=[
+                    positive_input,
+                    input_image,
+                    neg_input,
+                    seed_input,
+                    steps_input,
+                    style_strength_ratio_input,
+                ],
+                outputs=output_image,
+            )
+            gr.Examples(
+                [
+                    [prompt, negative_prompt],
+                    [
+                        "A woman img wearing a Christmas hat",
+                        negative_prompt,
+                    ],
+                    [
+                        "A man img in a helmet and vest riding a motorcycle",
+                        negative_prompt,
+                    ],
+                    [
+                        "photo of a middle-aged man img sitting on a plush leather couch, and watching television show",
+                        negative_prompt,
+                    ],
+                    [
+                        "photo of a skilled doctor img in a pristine white lab coat enjoying a delicious meal in a sophisticated dining room",
+                        negative_prompt,
+                    ],
+                    [
+                        "photo of superman img flying through a vibrant sunset sky, with his cape billowing in the wind",
+                        negative_prompt,
+                    ],
+                ],
+                [positive_input, neg_input],
+            )
+    return demo

notebooks/photo-maker/photo-maker.ipynb

@@ -975,15 +975,10 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "38788603-90b8-40e1-9bee-6743a0175417",
-   "metadata": {
-    "scrolled": true
-   },
+   "id": "ae437050",
+   "metadata": {},
    "outputs": [],
    "source": [
-    "import gradio as gr\n",
-    "\n",
-    "\n",
     "def generate_from_text(text_promt, input_image, neg_prompt, seed, num_steps, style_strength_ratio):\n",
     "    \"\"\"\n",
     "    Helper function for generating result image from prompt text\n",
@@ -1014,62 +1009,30 @@
     "        width=1024,\n",
     "    ).images[0]\n",
     "\n",
-    "    return result\n",
-    "\n",
-    "\n",
-    "with gr.Blocks() as demo:\n",
-    "    with gr.Column():\n",
-    "        with gr.Row():\n",
-    "            input_image = gr.Image(label=\"Your image\", sources=[\"upload\"], type=\"pil\")\n",
-    "            output_image = gr.Image(label=\"Generated Images\", type=\"pil\")\n",
-    "        positive_input = gr.Textbox(label=f\"Text prompt, Trigger words is '{trigger_word}'\")\n",
-    "        neg_input = gr.Textbox(label=\"Negative prompt\")\n",
-    "        with gr.Row():\n",
-    "            seed_input = gr.Slider(0, 10_000_000, value=42, label=\"Seed\")\n",
-    "            steps_input = gr.Slider(label=\"Steps\", value=10, minimum=5, maximum=50, step=1)\n",
-    "            style_strength_ratio_input = gr.Slider(label=\"Style strength ratio\", value=20, minimum=5, maximum=100, step=5)\n",
-    "            btn = gr.Button()\n",
-    "        btn.click(\n",
-    "            generate_from_text,\n",
-    "            [\n",
-    "                positive_input,\n",
-    "                input_image,\n",
-    "                neg_input,\n",
-    "                seed_input,\n",
-    "                steps_input,\n",
-    "                style_strength_ratio_input,\n",
-    "            ],\n",
-    "            output_image,\n",
-    "        )\n",
-    "        gr.Examples(\n",
-    "            [\n",
-    "                [prompt, negative_prompt],\n",
-    "                [\n",
-    "                    \"A woman img wearing a Christmas hat\",\n",
-    "                    negative_prompt,\n",
-    "                ],\n",
-    "                [\n",
-    "                    \"A man img in a helmet and vest riding a motorcycle\",\n",
-    "                    negative_prompt,\n",
-    "                ],\n",
-    "                [\n",
-    "                    \"photo of a middle-aged man img sitting on a plush leather couch, and watching television show\",\n",
-    "                    negative_prompt,\n",
-    "                ],\n",
-    "                [\n",
-    "                    \"photo of a skilled doctor img in a pristine white lab coat enjoying a delicious meal in a sophisticated dining room\",\n",
-    "                    negative_prompt,\n",
-    "                ],\n",
-    "                [\n",
-    "                    \"photo of superman img flying through a vibrant sunset sky, with his cape billowing in the wind\",\n",
-    "                    negative_prompt,\n",
-    "                ],\n",
-    "            ],\n",
-    "            [positive_input, neg_input],\n",
-    "        )\n",
+    "    return result"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "38788603-90b8-40e1-9bee-6743a0175417",
+   "metadata": {
+    "scrolled": true
+   },
+   "outputs": [],
+   "source": [
+    "if not Path(\"gradio_helper.py\").exists():\n",
+    "    r = requests.get(url=\"https://raw.githubusercontent.com/openvinotoolkit/openvino_notebooks/latest/notebooks/photo-maker/gradio_helper.py\")\n",
+    "    open(\"gradio_helper.py\", \"w\").write(r.text)\n",
+    "\n",
+    "from gradio_helper import make_demo\n",
     "\n",
+    "demo = make_demo(fn=generate_from_text)\n",
     "\n",
-    "demo.queue().launch()\n",
+    "try:\n",
+    "    demo.queue().launch(debug=True)\n",
+    "except Exception:\n",
+    "    demo.queue().launch(debug=True, share=True)\n",
     "# if you are launching remotely, specify server_name and server_port\n",
     "# demo.launch(server_name='your server name', server_port='server port in int')\n",
     "# Read more in the docs: https://gradio.app/docs/"
@@ -1084,7 +1047,8 @@
    },
    "outputs": [],
    "source": [
-    "demo.close()"
+    "# please uncomment and run this cell for stopping gradio interface\n",
+    "# demo.close()"
    ]
   }
  ],

notebooks/pix2struct-docvqa/gradio_helper.py

@@ -0,0 +1,39 @@
+from pathlib import Path
+from typing import Callable
+import gradio as gr
+import requests
+from PIL import Image
+
+
+example_images_urls = [
+    "https://github.com/openvinotoolkit/openvino_notebooks/assets/29454499/94ef687c-aebb-452b-93fe-c7f29ce19503",
+    "https://github.com/openvinotoolkit/openvino_notebooks/assets/29454499/70b2271c-9295-493b-8a5c-2f2027dcb653",
+    "https://github.com/openvinotoolkit/openvino_notebooks/assets/29454499/1e2be134-0d45-4878-8e6c-08cfc9c8ea3d",
+]
+
+file_names = ["eiffel_tower.png", "exsibition.jpeg", "population_table.jpeg"]
+
+for img_url, image_file in zip(example_images_urls, file_names):
+    if not Path(image_file).exists():
+        Image.open(requests.get(img_url, stream=True).raw).save(image_file)
+
+questions = [
+    "What is Eiffel tower tall?",
+    "When is the coffee break?",
+    "What the population of Stoddard?",
+]
+
+examples = [list(pair) for pair in zip(file_names, questions)]
+
+
+def make_demo(fn: Callable):
+    demo = gr.Interface(
+        fn=fn,
+        inputs=["image", "text"],
+        outputs="text",
+        title="Pix2Struct for DocVQA",
+        examples=examples,
+        cache_examples=False,
+        allow_flagging="never",
+    )
+    return demo

notebooks/pix2struct-docvqa/pix2struct-docvqa.ipynb

@@ -303,43 +303,19 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "import gradio as gr\n",
-    "\n",
-    "example_images_urls = [\n",
-    "    \"https://github.com/openvinotoolkit/openvino_notebooks/assets/29454499/94ef687c-aebb-452b-93fe-c7f29ce19503\",\n",
-    "    \"https://github.com/openvinotoolkit/openvino_notebooks/assets/29454499/70b2271c-9295-493b-8a5c-2f2027dcb653\",\n",
-    "    \"https://github.com/openvinotoolkit/openvino_notebooks/assets/29454499/1e2be134-0d45-4878-8e6c-08cfc9c8ea3d\",\n",
-    "]\n",
-    "\n",
-    "file_names = [\"eiffel_tower.png\", \"exsibition.jpeg\", \"population_table.jpeg\"]\n",
-    "\n",
-    "for img_url, image_file in zip(example_images_urls, file_names):\n",
-    "    load_image(img_url).save(image_file)\n",
-    "\n",
-    "questions = [\n",
-    "    \"What is Eiffel tower tall?\",\n",
-    "    \"When is the coffee break?\",\n",
-    "    \"What the population of Stoddard?\",\n",
-    "]\n",
-    "\n",
-    "examples = [list(pair) for pair in zip(file_names, questions)]\n",
-    "\n",
-    "\n",
     "def generate(img, question):\n",
     "    inputs = processor(images=img, text=question, return_tensors=\"pt\")\n",
     "    predictions = ov_model.generate(**inputs, max_new_tokens=256)\n",
     "    return processor.decode(predictions[0], skip_special_tokens=True)\n",
     "\n",
     "\n",
-    "demo = gr.Interface(\n",
-    "    fn=generate,\n",
-    "    inputs=[\"image\", \"text\"],\n",
-    "    outputs=\"text\",\n",
-    "    title=\"Pix2Struct for DocVQA\",\n",
-    "    examples=examples,\n",
-    "    cache_examples=False,\n",
-    "    allow_flagging=\"never\",\n",
-    ")\n",
+    "if not Path(\"gradio_helper.py\").exists():\n",
+    "    r = requests.get(url=\"https://raw.githubusercontent.com/openvinotoolkit/openvino_notebooks/latest/notebooks/pix2struct-docvqa/gradio_helper.py\")\n",
+    "    open(\"gradio_helper.py\", \"w\").write(r.text)\n",
+    "\n",
+    "from gradio_helper import make_demo\n",
+    "\n",
+    "demo = make_demo(fn=generate)\n",
     "\n",
     "try:\n",
     "    demo.queue().launch(debug=True)\n",
@@ -349,6 +325,17 @@
     "# demo.launch(server_name='your server name', server_port='server port in int')\n",
     "# Read more in the docs: https://gradio.app/docs/"
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "582a1703",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# please uncomment and run this cell for stopping gradio interface\n",
+    "# demo.close()"
+   ]
   }
  ],
  "metadata": {

notebooks/pixart/gradio_helper.py

@@ -0,0 +1,31 @@
+from typing import Callable
+import gradio as gr
+import numpy as np
+
+MAX_SEED = np.iinfo(np.int32).max
+
+examples = [
+    ["A small cactus with a happy face in the Sahara desert.", 42],
+    ["an astronaut sitting in a diner, eating fries, cinematic, analog film", 42],
+    [
+        "Pirate ship trapped in a cosmic maelstrom nebula, rendered in cosmic beach whirlpool engine, volumetric lighting, spectacular, ambient lights, light pollution, cinematic atmosphere, art nouveau style, illustration art artwork by SenseiJaye, intricate detail.",
+        0,
+    ],
+    ["professional portrait photo of an anthropomorphic cat wearing fancy gentleman hat and jacket walking in autumn forest.", 0],
+]
+
+
+def make_demo(fn: Callable):
+    demo = gr.Interface(
+        fn=fn,
+        inputs=[
+            gr.Textbox(label="Caption"),
+            gr.Slider(0, MAX_SEED, label="Seed"),
+            gr.Textbox(label="Negative prompt"),
+            gr.Slider(2, 20, step=1, label="Number of inference steps", value=4),
+        ],
+        outputs="image",
+        examples=examples,
+        allow_flagging="never",
+    )
+    return demo

notebooks/pixart/pixart.ipynb

@@ -131,10 +131,7 @@
    "outputs": [],
    "source": [
     "from pathlib import Path\n",
-    "\n",
-    "import numpy as np\n",
     "import torch\n",
-    "\n",
     "import openvino as ov\n",
     "\n",
     "\n",
@@ -529,35 +526,20 @@
    },
    "outputs": [],
    "source": [
-    "import gradio as gr\n",
-    "\n",
-    "\n",
     "def generate(prompt, seed, negative_prompt, num_inference_steps):\n",
     "    generator = torch.Generator().manual_seed(seed)\n",
     "    image = pipe(prompt=prompt, negative_prompt=negative_prompt, num_inference_steps=num_inference_steps, generator=generator, guidance_scale=0.0).images[0]\n",
     "    return image\n",
     "\n",
     "\n",
-    "demo = gr.Interface(\n",
-    "    generate,\n",
-    "    [\n",
-    "        gr.Textbox(label=\"Caption\"),\n",
-    "        gr.Slider(0, np.iinfo(np.int32).max, label=\"Seed\"),\n",
-    "        gr.Textbox(label=\"Negative prompt\"),\n",
-    "        gr.Slider(2, 20, step=1, label=\"Number of inference steps\", value=4),\n",
-    "    ],\n",
-    "    \"image\",\n",
-    "    examples=[\n",
-    "        [\"A small cactus with a happy face in the Sahara desert.\", 42],\n",
-    "        [\"an astronaut sitting in a diner, eating fries, cinematic, analog film\", 42],\n",
-    "        [\n",
-    "            \"Pirate ship trapped in a cosmic maelstrom nebula, rendered in cosmic beach whirlpool engine, volumetric lighting, spectacular, ambient lights, light pollution, cinematic atmosphere, art nouveau style, illustration art artwork by SenseiJaye, intricate detail.\",\n",
-    "            0,\n",
-    "        ],\n",
-    "        [\"professional portrait photo of an anthropomorphic cat wearing fancy gentleman hat and jacket walking in autumn forest.\", 0],\n",
-    "    ],\n",
-    "    allow_flagging=\"never\",\n",
-    ")\n",
+    "if not Path(\"gradio_helper.py\").exists():\n",
+    "    r = requests.get(url=\"https://raw.githubusercontent.com/openvinotoolkit/openvino_notebooks/latest/notebooks/pixart/gradio_helper.py\")\n",
+    "    open(\"gradio_helper.py\", \"w\").write(r.text)\n",
+    "\n",
+    "from gradio_helper import make_demo\n",
+    "\n",
+    "demo = make_demo(fn=generate)\n",
+    "\n",
     "try:\n",
     "    demo.queue().launch(debug=True)\n",
     "except Exception:\n",
@@ -566,6 +548,17 @@
     "# demo.launch(server_name='your server name', server_port='server port in int')\n",
     "# Read more in the docs: https://gradio.app/docs/"
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "1d4e6624",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# please uncomment and run this cell for stopping gradio interface\n",
+    "# demo.close()"
+   ]
   }
  ],
  "metadata": {