[pinecone/vectordb] pinecone 튜토리얼

API_KEY

API_KEY=$PINECONE_API

Region

environment='gcp-starter'

import pinecone

API_KEY = 'your_api_key'
ENVIRONMENT = 'your environment'

#클라이언트 생성
pinecone.init(api_key=API_KEY, environment=ENVIRONMENT)

if index not in pinecone.list_indexes():
	index = pinecone.create_index('tutorial', dimension=1536, metric='cosine')
    
index = pinecone.Index('tutorial')

df = pd.DataFrame(data={'question':question_list, 'vector':vector_list})

index.upsert(vectors=zip(df.question, df.vector))
벡터를 생성하는 데 사용할 임베딩 모델은 text-embedding-ada-0021536차원의 벡터를 출력합니다. ~~텍스트~~