Open AI:
Knowledge Retrieval:
Once a file is uploaded and passed to the Assistant, OpenAI will automatically chunk your documents, index and store the embeddings, and implement vector search to retrieve relevant content to answer user queries.
it works on vector search which will consider the most similar vector to the query, due to which the response will be based on that one vector,
is there any way so that the whole file is considered and the response is based on the whole file?
Example :
I have uploaded the file (contains information of 20 hotels ) which is 15000 words. when i asked the assistant ( provide me names of all the hotels).it only returned the first 2 hotel names.
I want it to return names of 20 hotels