Recreate the storage context

Amardeep · 2023-07-10T16:52:49.675Z

I ran across a problem with the following code: llama_index import StorageContext, load_index_from_storageRecreate the storage context StorageContext.from_defaults(persist_dir='./storage') storage_contextindex load = load_index_from_storage(storage_context)When I load a new document to generate a vector index, it creates a new vector index file, but I want to build a vector index in append mode, which stores my vector from the previous file as well as creates a vector index for the new file.can some body help me with it

You can append new files by adding Document object to the existing Index.

Code structure would look something like this

Plain Text

service_context = ServiceContext.from_defaults()
doc = Document(text="this is a document lol!")
new_index = GPTVectorStoreIndex.from_documents([doc], service_context=service_context)
new_index.as_query_engine().query("hello world")
new_index.storage_context.persist()

# creating a new doc 
doc = Document(text="this is awesome news")
# appending the new doc to the existing index
new_index.insert(doc)
# persisting the new changes into the storage folder
new_index.storage_context.persist()

Find answers from the community

Recreate the storage context

Recreate the storage context