Cetera chatbot prototipe

Upload a Document

Upload a file to the specified Milvus collection

Select file

Collection Name

Type the name of a new collection or select an existing one

Make a Query

Select a Milvus collection and ask a question

Collection

Your Question

Embedding Model Select the embedding model used to convert text into numerical vectors. The resulting vectors can the be used in similarity search queries to find relevant content

Context Window 1 Specifies how many sentences immediately preceding and following each text chunk should be included, in order to provide richer context and improve the relevance of the retrieved information

Similarity Threshold 95 Percentile threshold used to detect natural breakpoints in the text. A higher value means that only the largest semantic distances between sentences are considered breakpoints, resulting in fewer split points and larger text chunks. A lower value considers more distances as breakpoints, creating more, smaller chunks

Min Chunk Size 3 Sets the minimum number of sentences that each text chunk should contain. This ensures that chunks are not too short, preserving enough context for meaningful analysis or search

Upload a Document

Make a Query

LLM Answer

Chunks retrieved from Milvus