#Summary of documentation

1 messages · Page 1 of 1 (latest)

rocky marsh
#

Please provide me with a summary of any part of your documentation.

lunar lakeBOT
#

Hi @rocky marsh. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:

#

Sure, here are some key points from the documentation and guides provided by OpenAI:

  1. What makes documentation good: Good documentation should be broadly helpful, written simply, and avoid abbreviations. It should offer solutions to potential problems, use specific and accurate terminology, and keep code examples general and exportable. Topics should be prioritized by value, bad habits should not be taught, and topics should be introduced with a broad opening. Source (https://github.com/openai/openai-cookbook/blob/main/articles/what_makes_documentation_good.md)

  2. How to evaluate a summarization task: For evaluating a summarization task, two generated summaries are compared with a reference human-written summary. Evaluation metrics like ROUGE and BERTScore are used for this purpose. Source (https://github.com/openai/openai-cookbook/blob/main/examples/evaluation/How_to_eval_abstractive_summarization.ipynb)

  3. Prompt engineering: To summarize a very long document, a sequence of queries can be used to summarize each section of the document. Section summaries can be concatenated and summarized, producing summaries of summaries. This process can proceed recursively until an entire document is summarized. Source (https://platform.openai.com/docs/guides/prompt-engineering)

  4. How to use functions with a knowledge base: A function can be used to summarize key points extracted from an academic paper. The summary should highlight the core argument, conclusions, and evidence, and answer the user's query. Source (https://github.com/openai/openai-cookbook/blob/main/examples/How_to_call_functions_for_knowledge_retrieval.ipynb)

#

Please note that this is a high-level summary and for detailed understanding, you should refer to the original documentation.

Note: you can ask me a follow up question by @ mentioning me again :speech_balloon:

#

@rocky marsh kapa.ai is still learning and improving, please let us know how it did by reacting below

rocky marsh
#

@lunar lake can you elaborate more in detail on number 3?

lunar lakeBOT
#

Let me check the knowledge base for you. One moment! :books:

#

Sure, I'd be happy to elaborate on point 3, which is about the strategy of summarizing long documents using OpenAI's language models.

Due to the fixed context length of models, they cannot be used to summarize a text longer than the context length minus the length of the generated summary in a single query. To overcome this limitation, a sequence of queries can be used to summarize each section of the document. These section summaries can then be concatenated and summarized, producing summaries of summaries. This process can proceed recursively until an entire document is summarized.

If it’s necessary to use information about earlier sections in order to make sense of later sections, then a further trick that can be useful is to include a running summary of the text that precedes any given point in the book while summarizing content at that point.

The effectiveness of this procedure for summarizing books has been studied in previous research by OpenAI using variants of GPT-3. You can find more about this in the Prompt Engineering Guide (https://platform.openai.com/docs/guides/prompt-engineering) under the section "Strategy: Write clear instructions > Tactic: Summarize long documents piecewise and construct a full summary recursively".

Note: you can ask me a follow up question by @ mentioning me again :speech_balloon: