hard20 pts · first answer countsprompt-engineeringcontexttokens
How much room is left for documents
A model has a 128,000-token context window. Your system prompt is 6,000 tokens, chat history is capped at 10,000 tokens, and you reserve 4,000 tokens for the reply. Each retrieved chunk is 800 tokens. What is the maximum number of whole chunks you can add?
Discussion 0
Join free to post in the discussion.
Related questions
- What a token isPrompt Engineer
- Same text, more tokensPrompt Engineer
- What you pay forPrompt Engineer
- What top_p controlsPrompt Engineer
- What a stop sequence doesPrompt Engineer
- Reply ends mid-sentencePrompt Engineer
- Estimating tokens from wordsPrompt Engineer
- Letting the model say it does not knowPrompt Engineer