python 操作向量数据库qdrant

from sentence_transformers import SentenceTransformer

#模型已下载到本地
MODELBGE = SentenceTransformer(BASE_DIR + '/bge-large-zh')
class LocalEmbedding():
    def __init__(self, text, data=None):
        self.text = text#text必须是一个list：['文本1','文本2']
        self.data = data#list 
　　#查询时：搜索关键词+官方给定的instruction
    def achieve_emb(self):
        try:
            MODELBGE = SentenceTransformer(BASE_DIR + '/bge-large-zh')
            instruction = "为这个句子生成表示以用于检索相关文章："
            q_embeddings = MODELBGE.encode([instruction + self.text], normalize_embeddings=True).tolist()
            self.data.extend(q_embeddings)
            return self.data
        except Exception as e:
            print('本地embedding错误', e)
            return self.data
　　#入库文本的向量化
    def achieve_emb_txt(self):
        try:
            p_embeddings = MODELBGE.encode(self.text, normalize_embeddings=True).tolist()
            return p_embeddings
        except Exception as e:
            print('本地embedding错误', e)
            return self.data

def save_qdrant_url(i_path,file_name):
　　 '''
    i_path:文件路径
    file_name:文件名
    '''
    try:
　　　　 #连接qdrant数据库
        client = QdrantClient('ip', port=6333)
        collection_name = '自定义的集合名称'
        try:
            table_collection = client.get_collection(collection_name=f"{collection_name}")
            print(table_collection)
            print(f'集合已存在:{collection_name}')
            #update_collection更新集合，不会覆盖原有集合内的数据。
            client.update_collection(
                collection_name=collection_name
            )
        except:
            print(f'创建新集合：{collection_name}')
            #recreate_collection创建新的集合，原有数据清空
            client.recreate_collection(
                collection_name=collection_name,
                vectors_config=VectorParams(size=1024, distance=Distance.COSINE),
            )
        print(f'正在执行json文件为{file_name}')
        with open(f'{i_path}', 'r', encoding='utf-8') as f:
            data_lines = json.loads(f.read())
            text_all_list=data_lines[file_name]
            text_list = [text['content'] for text in text_all_list]

            tt = LocalEmbedding(text_list)
            vector_list = tt.achieve_emb_txt()
　　　　　　　#upload_collection批量存储，注意向量化的列表长度vectors与文本内容列表长度payload要一一对应。
            client.upload_collection(
                collection_name=collection_name,
                vectors=vector_list,
                payload=text_all_list,
                ids=None,
                batch_size=256,
                parallel=2
            )

            print(f'文件：{file_name} qdrant数据存储成功')
    except Exception as e:
        print(f'save_qdrant_data存储失败文件为:{file_name},失败原因为{e}')