-
Notifications
You must be signed in to change notification settings - Fork 0
/
summary_data_to_pinecone.py
95 lines (78 loc) · 3.23 KB
/
summary_data_to_pinecone.py
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
import os
import json
from dotenv import load_dotenv
import torch
from transformers import AutoModel, AutoTokenizer
import pinecone
import hashlib
import numpy as np
# .env 파일에서 환경 변수 로드
load_dotenv()
# Pinecone API 키 설정
PINECONE_API_KEY = os.getenv("PINECONE_API_KEY")
# Pinecone 초기화
pinecone.init(api_key=PINECONE_API_KEY, environment='us-west1-gcp')
# 인덱스 이름과 설정
index_name = "legal-docs"
dimension = 768 # HuggingFace KoSimCSE 로버타 모델의 임베딩 차원
metric = "cosine"
# 인덱스 존재 여부 확인 및 생성
if index_name not in pinecone.list_indexes():
pinecone.create_index(
name=index_name,
dimension=dimension,
metric=metric,
pod_type="p1"
)
index = pinecone.Index(index_name)
class MyEmbeddingModel: # 커스텀 임베딩 클래스
def __init__(self, model_name): # 클래스 생성자
self.tokenizer = AutoTokenizer.from_pretrained(model_name)
self.model = AutoModel.from_pretrained(model_name)
def embed_documents(self, doc): # 문서 임베딩 메서드
inputs = self.tokenizer(doc, padding=True, truncation=True, return_tensors="pt", max_length=512)
with torch.no_grad():
outputs = self.model(**inputs)
embeddings = outputs.last_hidden_state.mean(dim=1).tolist()
return embeddings
def embed_query(self, text): # 쿼리 임베딩 메서드
inputs = self.tokenizer([text], padding=True, truncation=True, return_tensors="pt", max_length=512)
with torch.no_grad():
outputs = self.model(**inputs)
embedding = outputs.last_hidden_state.mean(dim=1).squeeze().tolist()
return embedding
embed_model_name = "BM-K/KoSimCSE-roberta-multitask"
embedding_model = MyEmbeddingModel(embed_model_name)
def generate_vector_id(text, index):
unique_string = f"{text}_{index}"
return hashlib.md5(unique_string.encode('utf-8')).hexdigest()
def insert_data_to_pinecone(data):
vectors = []
for item in data:
for i, detail in enumerate(item.get('세부내용', [])):
text = detail['내용'] if isinstance(detail['내용'], str) else ' '.join(detail['내용'])
inputs = embedding_model.tokenizer(text, return_tensors='pt')
with torch.no_grad():
embeddings = embedding_model.model(**inputs).last_hidden_state.mean(dim=1).squeeze().tolist()
vector_id = generate_vector_id(text, i)
vectors.append({
'id': vector_id,
'values': embeddings,
'metadata': {
'주제': item['주제'],
'항목': item['항목'],
'설명': detail['설명'],
'내용': detail['내용']
}
})
# Pinecone에 벡터 삽입
index.upsert(vectors)
if __name__ == "__main__":
# 병합된 JSON 파일 경로
json_file_path = 'datas/summary/건설일용근로자_생활법령.json'
# JSON 파일 읽기
with open(json_file_path, 'r', encoding='utf-8') as file:
data = json.load(file)
# 데이터 Pinecone에 삽입
insert_data_to_pinecone(data)
print("JSON data embedded and uploaded to Pinecone index.")