Upload sd_token_similarity_calculator.ipynb

sd_token_similarity_calculator.ipynb

@@ -82,10 +82,29 @@
         "mix_method = \"None\""
       ],
       "metadata": {
-        "id": "Ch9puvwKH1s3"
+        "id": "Ch9puvwKH1s3",
+        "collapsed": true,
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "outputId": "982a9210-a3fd-4d90-bef7-5aa6f5864797"
       },
-      "execution_count": null,
-      "outputs": []
+      "execution_count": 2,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "Cloning into 'sd_tokens'...\n",
+            "remote: Enumerating objects: 10, done.\u001b[K\n",
+            "remote: Counting objects: 100% (7/7), done.\u001b[K\n",
+            "remote: Compressing objects: 100% (7/7), done.\u001b[K\n",
+            "remote: Total 10 (delta 1), reused 0 (delta 0), pack-reused 3 (from 1)\u001b[K\n",
+            "Unpacking objects: 100% (10/10), 306.93 KiB | 4.72 MiB/s, done.\n",
+            "/content/sd_tokens\n"
+          ]
+        }
+      ]
     },
     {
       "cell_type": "code",
@@ -106,10 +125,22 @@
         "#You can leave the 'prompt' field empty to get a random value tensor. Since the tensor is random value, it will not correspond to any tensor in the vocab.json list , and this it will have no ID."
       ],
       "metadata": {
-        "id": "RPdkYzT2_X85"
+        "id": "RPdkYzT2_X85",
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "outputId": "86f2f01e-6a04-4292-cee7-70fd8398e07f"
       },
-      "execution_count": null,
-      "outputs": []
+      "execution_count": 3,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "[49406, 8922, 49407]\n"
+          ]
+        }
+      ]
     },
     {
       "cell_type": "code",
@@ -136,7 +167,7 @@
       "metadata": {
         "id": "YqdiF8DIz9Wu"
       },
-      "execution_count": null,
+      "execution_count": 4,
       "outputs": []
     },
     {
@@ -189,10 +220,24 @@
         "#OPTIONAL : Add/subtract + normalize above result with another token. Leave field empty to get a random value tensor"
       ],
       "metadata": {
-        "id": "oXbNSRSKPgRr"
+        "id": "oXbNSRSKPgRr",
+        "collapsed": true,
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "outputId": "76f8ec94-d29c-46d9-893b-49875f3a1949"
       },
-      "execution_count": null,
-      "outputs": []
+      "execution_count": 5,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "Tokenized prompt  'mix_with' tensor C is a random valued tensor with no ID\n",
+            "No operation\n"
+          ]
+        }
+      ]
     },
     {
       "cell_type": "code",
@@ -230,10 +275,30 @@
         "#Produce a list id IDs that are most similiar to the prompt ID at positiion 1 based on above result"
       ],
       "metadata": {
-        "id": "juxsvco9B0iV"
+        "id": "juxsvco9B0iV",
+        "collapsed": true,
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "outputId": "dc893bbf-e9cb-425c-95b8-ffafd3ab2fbc"
       },
-      "execution_count": null,
-      "outputs": []
+      "execution_count": 6,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "Calculated all cosine-similarities between the token banana</w> with Id_A = 8922 with the the rest of the 49407 tokens as a 1x49407 tensor\n"
+          ]
+        }
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "source": [],
+      "metadata": {
+        "id": "cYYu5C5C6MHH"
+      }
     },
     {
       "cell_type": "code",
@@ -260,10 +325,321 @@
       ],
       "metadata": {
         "id": "YIEmLAzbHeuo",
-        "collapsed": true
+        "collapsed": true,
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "outputId": "4a2fa70f-16ff-4bba-fb01-d39ad697d4ff"
       },
-      "execution_count": null,
-      "outputs": []
+      "execution_count": 7,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "banana</w>\n",
+            "similiarity = 100.0 %\n",
+            "--------\n",
+            "bananas</w>\n",
+            "similiarity = 38.93 %\n",
+            "--------\n",
+            "banan\n",
+            "similiarity = 30.8 %\n",
+            "--------\n",
+            "ðŁįĮ</w>\n",
+            "similiarity = 27.12 %\n",
+            "--------\n",
+            "pineapple</w>\n",
+            "similiarity = 19.7 %\n",
+            "--------\n",
+            "chicken</w>\n",
+            "similiarity = 19.24 %\n",
+            "--------\n",
+            "potassium</w>\n",
+            "similiarity = 19.21 %\n",
+            "--------\n",
+            "sausage</w>\n",
+            "similiarity = 19.07 %\n",
+            "--------\n",
+            "lemon</w>\n",
+            "similiarity = 18.82 %\n",
+            "--------\n",
+            "orange</w>\n",
+            "similiarity = 18.42 %\n",
+            "--------\n",
+            "peanut</w>\n",
+            "similiarity = 17.84 %\n",
+            "--------\n",
+            "parachute</w>\n",
+            "similiarity = 17.19 %\n",
+            "--------\n",
+            "duck\n",
+            "similiarity = 16.8 %\n",
+            "--------\n",
+            "yellow</w>\n",
+            "similiarity = 16.21 %\n",
+            "--------\n",
+            "grape</w>\n",
+            "similiarity = 16.19 %\n",
+            "--------\n",
+            "kangaroo</w>\n",
+            "similiarity = 16.13 %\n",
+            "--------\n",
+            "apple</w>\n",
+            "similiarity = 16.13 %\n",
+            "--------\n",
+            "tangerine</w>\n",
+            "similiarity = 16.08 %\n",
+            "--------\n",
+            "giraffe</w>\n",
+            "similiarity = 16.04 %\n",
+            "--------\n",
+            "mango</w>\n",
+            "similiarity = 16.03 %\n",
+            "--------\n",
+            "rubber</w>\n",
+            "similiarity = 15.95 %\n",
+            "--------\n",
+            "bamboo</w>\n",
+            "similiarity = 15.88 %\n",
+            "--------\n",
+            "umbrella</w>\n",
+            "similiarity = 15.82 %\n",
+            "--------\n",
+            "nutella</w>\n",
+            "similiarity = 15.69 %\n",
+            "--------\n",
+            "ferrari</w>\n",
+            "similiarity = 15.69 %\n",
+            "--------\n",
+            "oranges</w>\n",
+            "similiarity = 15.65 %\n",
+            "--------\n",
+            "peanuts</w>\n",
+            "similiarity = 15.62 %\n",
+            "--------\n",
+            "ali</w>\n",
+            "similiarity = 15.49 %\n",
+            "--------\n",
+            "cucumber</w>\n",
+            "similiarity = 15.32 %\n",
+            "--------\n",
+            "potato</w>\n",
+            "similiarity = 15.22 %\n",
+            "--------\n",
+            "monkey</w>\n",
+            "similiarity = 15.2 %\n",
+            "--------\n",
+            "croissant</w>\n",
+            "similiarity = 15.18 %\n",
+            "--------\n",
+            "papaya</w>\n",
+            "similiarity = 15.17 %\n",
+            "--------\n",
+            "christmas</w>\n",
+            "similiarity = 15.12 %\n",
+            "--------\n",
+            "sandwich</w>\n",
+            "similiarity = 15.0 %\n",
+            "--------\n",
+            "rainbow</w>\n",
+            "similiarity = 14.98 %\n",
+            "--------\n",
+            "tomato</w>\n",
+            "similiarity = 14.96 %\n",
+            "--------\n",
+            "martini</w>\n",
+            "similiarity = 14.93 %\n",
+            "--------\n",
+            "cabaret</w>\n",
+            "similiarity = 14.83 %\n",
+            "--------\n",
+            "ginger</w>\n",
+            "similiarity = 14.82 %\n",
+            "--------\n",
+            "animal</w>\n",
+            "similiarity = 14.76 %\n",
+            "--------\n",
+            "vanilla</w>\n",
+            "similiarity = 14.73 %\n",
+            "--------\n",
+            "mustache</w>\n",
+            "similiarity = 14.64 %\n",
+            "--------\n",
+            "lime</w>\n",
+            "similiarity = 14.62 %\n",
+            "--------\n",
+            "sickle</w>\n",
+            "similiarity = 14.6 %\n",
+            "--------\n",
+            "vista</w>\n",
+            "similiarity = 14.53 %\n",
+            "--------\n",
+            "coconut</w>\n",
+            "similiarity = 14.52 %\n",
+            "--------\n",
+            "kara</w>\n",
+            "similiarity = 14.46 %\n",
+            "--------\n",
+            "alligator</w>\n",
+            "similiarity = 14.39 %\n",
+            "--------\n",
+            "blueberry</w>\n",
+            "similiarity = 14.34 %\n",
+            "--------\n",
+            "squirrel</w>\n",
+            "similiarity = 14.29 %\n",
+            "--------\n",
+            "atore</w>\n",
+            "similiarity = 14.19 %\n",
+            "--------\n",
+            "watermelon</w>\n",
+            "similiarity = 14.13 %\n",
+            "--------\n",
+            "nana</w>\n",
+            "similiarity = 14.09 %\n",
+            "--------\n",
+            "latex</w>\n",
+            "similiarity = 14.08 %\n",
+            "--------\n",
+            "agricultural</w>\n",
+            "similiarity = 14.02 %\n",
+            "--------\n",
+            "zucchini</w>\n",
+            "similiarity = 14.0 %\n",
+            "--------\n",
+            "saxophone</w>\n",
+            "similiarity = 13.93 %\n",
+            "--------\n",
+            "mozzarella</w>\n",
+            "similiarity = 13.91 %\n",
+            "--------\n",
+            "eggplant</w>\n",
+            "similiarity = 13.9 %\n",
+            "--------\n",
+            "pickle</w>\n",
+            "similiarity = 13.89 %\n",
+            "--------\n",
+            "tortilla</w>\n",
+            "similiarity = 13.88 %\n",
+            "--------\n",
+            "maniac</w>\n",
+            "similiarity = 13.84 %\n",
+            "--------\n",
+            "milk</w>\n",
+            "similiarity = 13.83 %\n",
+            "--------\n",
+            "cellphone</w>\n",
+            "similiarity = 13.78 %\n",
+            "--------\n",
+            "duck</w>\n",
+            "similiarity = 13.73 %\n",
+            "--------\n",
+            "umbrel\n",
+            "similiarity = 13.71 %\n",
+            "--------\n",
+            "fanny</w>\n",
+            "similiarity = 13.69 %\n",
+            "--------\n",
+            "twister</w>\n",
+            "similiarity = 13.67 %\n",
+            "--------\n",
+            "moustache</w>\n",
+            "similiarity = 13.66 %\n",
+            "--------\n",
+            "manafort</w>\n",
+            "similiarity = 13.66 %\n",
+            "--------\n",
+            "grapefruit</w>\n",
+            "similiarity = 13.6 %\n",
+            "--------\n",
+            "broom</w>\n",
+            "similiarity = 13.59 %\n",
+            "--------\n",
+            "scorpion</w>\n",
+            "similiarity = 13.59 %\n",
+            "--------\n",
+            "fruit\n",
+            "similiarity = 13.57 %\n",
+            "--------\n",
+            "agan\n",
+            "similiarity = 13.53 %\n",
+            "--------\n",
+            "sunflower</w>\n",
+            "similiarity = 13.49 %\n",
+            "--------\n",
+            "banc\n",
+            "similiarity = 13.46 %\n",
+            "--------\n",
+            "literature</w>\n",
+            "similiarity = 13.45 %\n",
+            "--------\n",
+            "pelican</w>\n",
+            "similiarity = 13.43 %\n",
+            "--------\n",
+            "breakfast</w>\n",
+            "similiarity = 13.42 %\n",
+            "--------\n",
+            "pear</w>\n",
+            "similiarity = 13.42 %\n",
+            "--------\n",
+            "orange\n",
+            "similiarity = 13.4 %\n",
+            "--------\n",
+            "monet</w>\n",
+            "similiarity = 13.4 %\n",
+            "--------\n",
+            "snake</w>\n",
+            "similiarity = 13.32 %\n",
+            "--------\n",
+            "vampire</w>\n",
+            "similiarity = 13.32 %\n",
+            "--------\n",
+            "cinnamon</w>\n",
+            "similiarity = 13.3 %\n",
+            "--------\n",
+            "strawberries</w>\n",
+            "similiarity = 13.29 %\n",
+            "--------\n",
+            "butternut</w>\n",
+            "similiarity = 13.22 %\n",
+            "--------\n",
+            "sausages</w>\n",
+            "similiarity = 13.22 %\n",
+            "--------\n",
+            "iphone</w>\n",
+            "similiarity = 13.21 %\n",
+            "--------\n",
+            "egg\n",
+            "similiarity = 13.2 %\n",
+            "--------\n",
+            "capu\n",
+            "similiarity = 13.2 %\n",
+            "--------\n",
+            "mannequin</w>\n",
+            "similiarity = 13.19 %\n",
+            "--------\n",
+            "cucumbers</w>\n",
+            "similiarity = 13.16 %\n",
+            "--------\n",
+            "champagne</w>\n",
+            "similiarity = 13.15 %\n",
+            "--------\n",
+            "triangle</w>\n",
+            "similiarity = 13.14 %\n",
+            "--------\n",
+            "apples</w>\n",
+            "similiarity = 13.09 %\n",
+            "--------\n",
+            "dynamite</w>\n",
+            "similiarity = 13.08 %\n",
+            "--------\n",
+            "chocolate</w>\n",
+            "similiarity = 13.08 %\n",
+            "--------\n"
+          ]
+        }
+      ]
     },
     {
       "cell_type": "code",
@@ -280,10 +656,23 @@
         "#Valid ID ranges for id_for_token_A / id_for_token_B are between 0 and 49407"
       ],
       "metadata": {
-        "id": "MwmOdC9cNZty"
+        "id": "MwmOdC9cNZty",
+        "collapsed": true,
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "outputId": "0dd984d0-e253-4981-d72f-40aa83d57d8b"
       },
-      "execution_count": null,
-      "outputs": []
+      "execution_count": 8,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "The similarity between tokens A and B is 3.671 %\n"
+          ]
+        }
+      ]
     },
     {
       "cell_type": "code",
@@ -292,7 +681,7 @@
         "\n",
         "prompt_A = \"\" # @param {\"type\":\"string\",\"placeholder\":\"Write a prompt\"}\n",
         "prompt_B = \"\" # @param {\"type\":\"string\",\"placeholder\":\"Write a prompt\"}\n",
-        "use_token_padding = False # @param {type:\"boolean\"}\n",
+        "use_token_padding = True # @param {type:\"boolean\"}\n",
         "\n",
         "from transformers import  CLIPProcessor, CLIPModel\n",
         "\n",
@@ -307,7 +696,7 @@
         "ids_B = processor.tokenizer(text=prompt_B, padding=use_token_padding, return_tensors=\"pt\")\n",
         "text_encoding_B = model.get_text_features(**ids_B)\n",
         "\n",
-        "similarity_str =  'The similarity between the text_encoding for A and B is ' +  token_similarity(text_encoding_A[0] , text_encoding_B[0])\n",
+        "similarity_str =  'The similarity between the text_encoding for A:\"' + prompt_A + '\" and B: \"' + prompt_B +'\" is ' +  token_similarity(text_encoding_A[0] , text_encoding_B[0])\n",
         "\n",
         "\n",
         "print(similarity_str)\n",
@@ -319,10 +708,23 @@
         "\n"
       ],
       "metadata": {
-        "id": "QQOjh5BvnG8M"
+        "id": "QQOjh5BvnG8M",
+        "collapsed": true,
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "outputId": "8bd6eb94-c5a7-47e6-913b-346941b144a6"
       },
-      "execution_count": null,
-      "outputs": []
+      "execution_count": 11,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "The similarity between the text_encoding for A:\"one ripe banana\" and B: \"a long yellow fruit\" is 83.696 %\n"
+          ]
+        }
+      ]
     },
     {
       "cell_type": "markdown",