Find answers from the community

Home
Members
Torsten
T
Torsten
Offline, last seen 4 days ago
Joined September 25, 2024
Hello, I'm using the context mode chat engine and am testing out a new embedding model (OpenAI's text-embedding-3-large). However, when I then run my chat engine, I am no longer able to have an interactive conversation. Can someone help me out?
14 comments
T
W
L
Hi, my chat_engine (context mode) is currently quite slow. even on the first question. What are some good practices for increasing the efficiency of my chat engine?
8 comments
L
T
Hi, for my chat_engine I have set up my prompt like this: "You are a chatbot that answers questions about "company" models. \
Instructions:
  1. Never provide information that is not literally in the knowledge base.\
  2. If the answer cannot be found in the knowledge base tell the user that the\
    provided context does not tell you anything about the subject in question.\
    1. Always provide a link to the page that is used to generate the answer."
However, in some cases the chat_engine still uses information that is not in the context. Can someone help me with debugging?
2 comments
T
T
What happened to the documentation of older versions?
6 comments
L
T
v
Is it me or has gpt-4 become much more strict as of recently?
1 comment
T
Hi there, is there a way to set some sort of strictness for my context mode chat engine? i.e. how strict the engine is about returning information that is in the context or not?
4 comments
T
W
What are some methods to increase the speed of my chatengine?
3 comments
e
T
Hello, I get this error when using an index created with the text-embedding-3-large and gpt-4o as llm:
shapes (1536,) and (3072,) not aligned: 1536 (dim 0) != 3072 (dim 0)
How to fix this?
12 comments
T
W
What is currently the best combination of embedding model + gpt for a SimpleVectorStore and a context_mode chat_engine? text-embedding-3-small in combination with gpt-4o-2024-05-13 gives me terrible performance
20 comments
n
T
W
L
I'm in the process of updating my repo to a newer version. It seems like the LLMPredictor class has been depreciated. I used this to set up my ServiceContext. What should I use for that now? I cannot find it in the docs.
18 comments
T
L
T
Torsten
·

Prompt

Hi everyone, I have set up an index based on some documentation on a specific product of a company. What is the best way to make sure that the chat engine with chatmode "context" only uses information in the knowledge base to generate answers?
79 comments
W
T
Hello, does anyone have any tips for setting the parameters of the PromptHelper?
4 comments
T
L