adding text embedder sample for python

examples/text_embedder/python/text_embedder.ipynb

@@ -0,0 +1,213 @@
+{
+  "cells": [
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "h2q27gKz1H20"
+      },
+      "source": [
+        "**Copyright 2023 The MediaPipe Authors. All Rights Reserved.**"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {
+        "id": "TUfAcER1oUS6"
+      },
+      "outputs": [],
+      "source": [
+        "#@title Licensed under the Apache License, Version 2.0 (the \"License\");\n",
+        "# you may not use this file except in compliance with the License.\n",
+        "# You may obtain a copy of the License at\n",
+        "#\n",
+        "# https://www.apache.org/licenses/LICENSE-2.0\n",
+        "#\n",
+        "# Unless required by applicable law or agreed to in writing, software\n",
+        "# distributed under the License is distributed on an \"AS IS\" BASIS,\n",
+        "# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n",
+        "# See the License for the specific language governing permissions and\n",
+        "# limitations under the License."
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "L_cQX8dWu4Dv"
+      },
+      "source": [
+        "# Text Embedding with MediaPipe Tasks"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "source": [
+        "This notebook will show you how to use the MediaPipe Tasks Python API to compare text items, giving you a value for how similar they are. These values will range from -1 to 1 with 1 being the same text. This is done through cosine similarity."
+      ],
+      "metadata": {
+        "id": "jERPUr5-vbYk"
+      }
+    },
+    {
+      "cell_type": "markdown",
+      "source": [
+        "##Preparation\n",
+        "\n",
+        "Let's start with installing MediaPipe.\n",
+        "\n",
+        "* If you see an error about flatbuffers incompatibility, it's fine to ignore it.\n",
+        "MediaPipe requires a newer version of flatbuffers (v2), which is incompatible with the older version of Tensorflow (v2.9) currently preinstalled on Colab.\n",
+        "\n",
+        "* If you install MediaPipe outside of Colab, you only need to run pip install mediapipe. It isn't necessary to explicitly install flatbuffers."
+      ],
+      "metadata": {
+        "id": "7LSy09XEvuJy"
+      }
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 2,
+      "metadata": {
+        "id": "gxbHBsF-8Y_l",
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "outputId": "a6944bf4-3d62-4810-9b1a-339d953c9f69"
+      },
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "\u001b[31mERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.\n",
+            "tensorflow 2.9.2 requires flatbuffers<2,>=1.12, but you have flatbuffers 2.0 which is incompatible.\u001b[0m\u001b[31m\n",
+            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m33.0/33.0 MB\u001b[0m \u001b[31m13.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+            "\u001b[?25h"
+          ]
+        }
+      ],
+      "source": [
+        "!pip install -q flatbuffers==2.0.0\n",
+        "!pip install -q mediapipe==0.9.0.1"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "a49D7h4TVmru"
+      },
+      "source": [
+        "After installing your dependencies, you can download the text embedder model that will be used for this example."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 3,
+      "metadata": {
+        "id": "OMjuVQiDYJKF"
+      },
+      "outputs": [],
+      "source": [
+        "#@title Start downloading here.\n",
+        "!wget -O embedder.tflite -q https://storage.googleapis.com/mediapipe-assets/mobilebert_embedding_with_metadata.tflite"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "Iy4r2_ePylIa"
+      },
+      "source": [
+        "## Running inference\n",
+        "\n",
+        "To run inference using the MediaPipe Text Embedder task, you just need to initialize the `TextEmbedder` using the model that you downloaded earlier, and then use that `TextEmbedder` to compare two separate strings. You can edit the two strings that will be used on the side of this section where you see `first_text` and `second_text`."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 7,
+      "metadata": {
+        "id": "Yl_Oiye4mUuo",
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "outputId": "d26b3a4b-c4d4-42d8-8ff0-38074817f4af"
+      },
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "0.9332579993217945\n"
+          ]
+        }
+      ],
+      "source": [
+        "from mediapipe.tasks import python\n",
+        "from mediapipe.tasks.python.components import processors\n",
+        "from mediapipe.tasks.python import text\n",
+        "from scipy.spatial import distance\n",
+        "\n",
+        "# Create your base options with the model that was downloaded earlier\n",
+        "base_options = python.BaseOptions(model_asset_path='embedder.tflite')\n",
+        "\n",
+        "# Set your values for using normalization and quantization\n",
+        "l2_normalize = True #@param {type:\"boolean\"}\n",
+        "quantize = False #@param {type:\"boolean\"}\n",
+        "\n",
+        "# Create the final set of options for the Embedder\n",
+        "options = text.TextEmbedderOptions(\n",
+        "    base_options=base_options, l2_normalize=l2_normalize, quantize=quantize)\n",
+        "\n",
+        "with text.TextEmbedder.create_from_options(options) as embedder:\n",
+        "  # Retrieve the first and second sets of text that will be compared\n",
+        "  first_text = \"I'm feeling so good\" #@param {type:\"string\"}\n",
+        "  second_text = \"I'm okay I guess\" #@param {type:\"string\"}\n",
+        "\n",
+        "  # Convert both sets of text to embeddings\n",
+        "  first_embedding_result = embedder.embed(first_text)\n",
+        "  second_embedding_result = embedder.embed(second_text)\n",
+        "  first_embedding_result.embeddings[0]\n",
+        "\n",
+        "  # Retrieve the cosine similarity value from both sets of text, then take the\n",
+        "  # cosine of that value to receie a decimal similarity value.\n",
+        "  similarity = 1 - distance.cosine(first_embedding_result.embeddings[0].embedding.astype('float'),\n",
+        "                                   second_embedding_result.embeddings[0].embedding.astype('float'))\n",
+        "  print(similarity)"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [],
+      "metadata": {
+        "id": "Yn0qjHhVVlc_"
+      },
+      "execution_count": null,
+      "outputs": []
+    }
+  ],
+  "metadata": {
+    "colab": {
+      "provenance": []
+    },
+    "kernelspec": {
+      "display_name": "Python 3 (ipykernel)",
+      "language": "python",
+      "name": "python3"
+    },
+    "language_info": {
+      "codemirror_mode": {
+        "name": "ipython",
+        "version": 3
+      },
+      "file_extension": ".py",
+      "mimetype": "text/x-python",
+      "name": "python",
+      "nbconvert_exporter": "python",
+      "pygments_lexer": "ipython3",
+      "version": "3.7.13"
+    }
+  },
+  "nbformat": 4,
+  "nbformat_minor": 0
+}