I install 0.8.63.post2 but i still can't use "gpt-4-1106-preview" and "gpt-3.5-turbo-1106" =(((( Why?

Why do question marks, spaces, and periods affect the choice of llmaindex nodes?
For example, if I don't put a question mark at the end, the text fragment is selected correctly. If there is a question mark, then the fragment is not on the question, the same with a space, period, etc.

Hello, everyone. I am interested in this argument "similarity_top_k".

Is it possible to make it dynamic? That is, if only one paragraph matches the most - similarity_top_k == 1. If there are several really relevant ones, similarity_top_k == 2. etc.

Hello, everyone. Could someone explain to me how the system of searching for suitable "pieces of text" in a document works in GPTVectorStoreIndex?
And how to make this selection more accurate, because sometimes it finds inappropriate passages.

Hi guys! i try use gpt-4o but got error:
Unknown model 'gpt-4o'. Please provide a valid OpenAI model name in: gpt-4, gpt-4-32k, gpt-4-1106-preview, gpt-4-0125-preview, gpt-4-turbo-preview, gpt-4-vision-preview, gpt-4-0613, gpt-4-32k-0613, gpt-4-0314, gpt-4-32k-0314, gpt-3.5-turbo, gpt-3.5-turbo-16k, gpt-3.5-turbo-0125, gpt-3.5-turbo-1106, gpt-3.5-turbo-0613, gpt-3.5-turbo-16k-0613, gpt-3.5-turbo-0301, text-davinci-003, text-davinci-002, gpt-3.5-turbo-instruct, text-ada-001, text-babbage-001, text-curie-001, ada, babbage, curie, davinci, gpt-35-turbo-16k, gpt-35-turbo, gpt-35-turbo-0125, gpt-35-turbo-1106, gpt-35-turbo-0613, gpt-35-turbo-16k-0613

I am currently using this type of vector storage:
index = GPTVectorStoreIndex.from_documents(
documents, service_context=service_context
)
index.storage_context.persist(persist_dir=index_name)

But when I upload files of about 40 MB, everything is indexed for a very long time and the response takes up to 2-3 minutes. Can Weaviate solve this problem?

Hello friends. On my project, the version of llamaindex =
0.8.66. And the processing of one request takes about 26 seconds. Is it faster on the new version of llamain? Just 26 seconds is too long...

Hello friends. I'm wondering if it is now possible to set a separation character for LLM other than "chunk_overlap", "chunk_size". I saw in LlamaIndex cloud that there is a menu where you can set the text splitter. But I don't see it in the documentation. The only way I found is to reassign the "get_nodes_from_node" function from the llama_index/node_parser/node_utils.py file

@kapa.ai i want to gen questions to my nodes with metadata "questions_this_excerpt_can_answer"

Hello Guys. I'm facing a problem with profitability.

Let me explain in more detail. I have a document that LlamaIndex works with. The document is 200+ pages long.
At every slightest change in the file, the document is re-indexed. Is there any way to index only the part of the document that has been updated. And update part of the LlamaIndex indexes?

Because when re-indexing a document of 200+ pages, it hits my wallet hard....

Hello, friends. I have created an assistant bot for one service.

It should communicate only in Ukrainian. But from time to time it can answer in English. How can I strictly limit it to use only Ukrainian?

Hi everyone, as you may already know, we have a new language - Mojo (a slightly "better" version of python). Mojo can easily use python modules.

So, I have a question for the LlamaIndex developers. Today we have a rather slow llamaindex (something like 20 sec). Maybe we can try to use Mojo to make it faster?

hi. Smb know analogue of LlamaIndex but for rust lang?