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
|
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"id": "a5caaba0-2c5d-4555-90eb-a7f8db555c2d",
"metadata": {},
"outputs": [],
"source": [
"import pandas as pd\n",
"import openai\n",
"from dotenv import load_dotenv, find_dotenv\n",
"import os\n",
"import json\n",
"from tqdm import tqdm\n"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "c17b66f4-c35b-4751-8dbd-44b827a81a39",
"metadata": {},
"outputs": [],
"source": [
"cypher_rag = pd.read_csv('../data/results/cypher_rag_mcq_output.csv')\n"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "4e8372a7-d61b-4b27-8161-3ab084c091e8",
"metadata": {},
"outputs": [],
"source": [
"llm_for_evaluation = 'gpt-4'\n"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "3930c00b-4647-4444-a2b6-f6a1030bd388",
"metadata": {},
"outputs": [],
"source": [
"load_dotenv(os.path.join(os.path.expanduser('~'), '.gpt_config.env'))\n",
"api_key = os.environ.get('API_KEY')\n",
"api_version = os.environ.get('API_VERSION')\n",
"resource_endpoint = os.environ.get('RESOURCE_ENDPOINT')\n",
"openai.api_type = 'azure'\n",
"openai.api_key = api_key\n",
"if resource_endpoint:\n",
" openai.api_base = resource_endpoint\n",
"if api_version:\n",
" openai.api_version = api_version"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "77dbdd34-f3a5-40b4-b0c7-1b6d2d6c5835",
"metadata": {},
"outputs": [],
"source": [
"def fetch_GPT_response(instruction, system_prompt, chat_model_id, chat_deployment_id, temperature=0):\n",
" # print('Calling OpenAI...')\n",
" response = openai.ChatCompletion.create(\n",
" temperature=temperature,\n",
" deployment_id=chat_deployment_id,\n",
" model=chat_model_id,\n",
" messages=[\n",
" {\"role\": \"system\", \"content\": system_prompt},\n",
" {\"role\": \"user\", \"content\": instruction}\n",
" ]\n",
" )\n",
" if 'choices' in response \\\n",
" and isinstance(response['choices'], list) \\\n",
" and len(response) >= 0 \\\n",
" and 'message' in response['choices'][0] \\\n",
" and 'content' in response['choices'][0]['message']:\n",
" return response['choices'][0]['message']['content']\n",
" else:\n",
" return 'Unexpected response'\n",
"\n",
"def get_GPT_response(instruction, system_prompt, chat_model_id, chat_deployment_id, temperature=0):\n",
" return fetch_GPT_response(instruction, system_prompt, chat_model_id, chat_deployment_id, temperature)\n",
" "
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "8285a1c7-4654-4728-b0d0-2a1300af26df",
"metadata": {},
"outputs": [],
"source": [
"system_prompt = '''\n",
"You are an expert evaluator of mcq question. You will be provided with an MCQ question which has 5 options out of which one option is correct.\n",
"You will also be provided with the correct label.\n",
"You will then provided with an answer.\n",
"You need to verify if that answer is correct or not, by checking it with the question and the label.\n",
"If the answer is correct, you return True, if the answer is not correct, you return False.\n",
"Return your answer in JSON format as follows:\n",
"{\"response\": <\"True\"/\"False\">}\n",
"\n",
"Example 1:\n",
"Question: Out of the given list, which Gene is associated with psoriasis and Takayasu's arteritis. Given list is: SHTN1, HLA-B, SLC14A2, BTBD9, DTNB\n",
"Label: HLA-B\n",
"Answer: The gene HLA-B is associated with both psoriasis and Takayasu's arteritis.\n",
"Your response: {\"response\": \"True\", \"reason\": \"Answer is in agreement with the label.\"}\n",
"\n",
"Example 2:\n",
"Question: Out of the given list, which Gene is associated with psoriasis and myelodysplastic syndrome. Given list is: NOD2, CHEK2, HLA-B, GCKR, PKNOX2\n",
"Label: HLA-B\n",
"Answer: The genes associated with psoriasis and myelodysplastic syndrome from the given list are HLA-B and CHEK2.\n",
"Your response: {\"response\": \"False\", \"reason\":\"Answer returns two options HLA-B and CHEK2, but the label says the correct answer is only HLA-B\"}\n",
"'''"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "76dfe699-2678-442c-b005-ec76a5989d69",
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"306it [11:29, 2.25s/it]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"CPU times: user 4.33 s, sys: 473 ms, total: 4.81 s\n",
"Wall time: 11min 29s\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"\n"
]
}
],
"source": [
"%%time\n",
"\n",
"cypher_rag_eval = []\n",
"for index, row in tqdm(cypher_rag.iterrows()):\n",
" question = row.text\n",
" label = row.label\n",
" answer = row.cypher_rag_answer\n",
" prompt = f'''\n",
" Question: {question}\n",
" Label: {label}\n",
" Answer: {answer}\n",
" '''\n",
" out = json.loads(get_GPT_response(prompt, system_prompt, llm_for_evaluation, llm_for_evaluation))\n",
" cypher_rag_eval.append(out[\"response\"])\n",
" "
]
},
{
"cell_type": "code",
"execution_count": 15,
"id": "c788084a-42df-48ab-9b3f-8feeef320a5a",
"metadata": {},
"outputs": [],
"source": [
"cypher_rag['cypher_rag_eval'] = cypher_rag_eval"
]
},
{
"cell_type": "code",
"execution_count": 18,
"id": "afbdf643-59b7-4af1-bf95-fe032f0dc4d9",
"metadata": {},
"outputs": [],
"source": [
"cypher_rag.to_csv('../data/results/cypher_rag_mcq_output_with_eval.csv', index=False)\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "86bcafaf-3945-4d63-a0a9-227736d7c672",
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python (cypher_rag)",
"language": "python",
"name": "cypher_rag"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.10.9"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
|
