File size: 5,336 Bytes
e42e9ae |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 |
{
"cells": [
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"# Evaluate Embedding Similarity Metrics"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"import openai, numpy as np"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"def cosine_distance(a, b):\n",
" \"\"\"Calculate the cosine distance between two numpy arrays.\n",
" \n",
" Parameters:\n",
" a (numpy array): First input array.\n",
" b (numpy array): Second input array.\n",
" \n",
" Returns:\n",
" float: Cosine distance between a and b.\n",
" \"\"\"\n",
" # Calculate dot product and magnitudes of the input arrays\n",
" dot = np.dot(a, b)\n",
" a_mag = np.linalg.norm(a)\n",
" b_mag = np.linalg.norm(b)\n",
" \n",
" if np.isclose(a_mag, 0, rtol=1e-9, atol=1e-12):\n",
" print(f\"a_mag is very small: {a_mag}\")\n",
" if np.isclose(b_mag, 0, rtol=1e-9, atol=1e-12):\n",
" print(f\"b_mag is very small: {b_mag}\")\n",
" \n",
" # Calculate and return the cosine distance\n",
" return 1.0 - (dot / (a_mag * b_mag))"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"def semantically_similar(string1, string2):\n",
" response = openai.Embedding.create(\n",
" input=[string1, string2],\n",
" engine=\"text-similarity-davinci-001\"\n",
" )\n",
" embedding_a = response['data'][0]['embedding']\n",
" embedding_b = response['data'][1]['embedding']\n",
" similarity_score = cosine_distance(embedding_a, embedding_b)\n",
" print(f\"similarity: {similarity_score}\")\n",
"\n",
" return similarity_score < 0.2"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"similarity: 0.22501948669661986\n",
"similarity: 0.2318907843871436\n",
"similarity: 0.12933868208210475\n",
"similarity: 0.10699853725782704\n"
]
},
{
"data": {
"text/plain": [
"(True,)"
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"semantically_similar(\"fight a war\", \"water supply\"),\n",
"semantically_similar(\"fight a war\", \"solar energy\"),\n",
"semantically_similar(\"fight a war\", \"defend a country\"),\n",
"semantically_similar(\"fight a war\", \"win a battle\"),"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"similarity: 0.2496415604648079\n"
]
},
{
"data": {
"text/plain": [
"False"
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"semantically_similar(\"the sky is blue\", \"I like to eat\")"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"similarity: 0.10193029028713485\n"
]
},
{
"data": {
"text/plain": [
"True"
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"semantically_similar(\"the cat meows\", \"the feline animal says\")"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"similarity: 0.19759407795526762\n"
]
},
{
"data": {
"text/plain": [
"True"
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"semantically_similar(\"what is the best way to win a war?\", \"strategizing a war\")"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"similarity: 0.1949772795717004\n"
]
},
{
"data": {
"text/plain": [
"True"
]
},
"execution_count": 9,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"semantically_similar(\"what is the best way to win a war?\", \"fight a war\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "chain",
"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": "3.9.16"
},
"orig_nbformat": 4
},
"nbformat": 4,
"nbformat_minor": 2
}
|