im trying to make a custom gpt that has access to some files I uploaded. but it refuses to look up the files for inference. any tips on how to make it always do retrieval/knowledge look up?
#custom gpt retrieval
low locust
clear quiverBOT
Hi @low locust. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:
Based on the information provided in the extracts, there are several steps you can take to ensure your custom GPT model has access to the files you uploaded and can use them for inference:
-
Upload the Files Correctly: Make sure you have uploaded the files correctly in the 'Knowledge' section of the GPT configuration tab. This allows you to provide additional context for your GPT to reference. Content from the files that are uploaded could be included in the output source (https://help.openai.com/en/articles/8554397-creating-a-gpt#h_9fdf0ceaf0).
-
Use Text Files: If you're uploading PDFs, it might be beneficial to convert them into .txt files instead. This could make it easier for the GPT model to understand and retrieve the information source (https://community.openai.com/t/501549).
-
Fine-Tuning: Fine-tuning the model with more information about the files might help, but it's not clear from the extracts whether this would increase the speed of retrieval source (https://community.openai.com/t/554222).
-
Check the File Size: If you're trying to upload a large amount of data or very large files, this could be causing issues. It might be worth trying to reduce the size of the files or the amount of data you're uploading source (https://community.openai.com/t/498177).
-
Internet Access: Remember that GPT models do not have access to the internet. Any access to other systems, including uploaded documents, has to be implemented by the services that drive the model source (https://community.openai.com/t/468615).
If you're still having issues, it might be worth reaching out to OpenAI support or the OpenAI community for further assistance.
Note: you can ask me a follow up question by @ mentioning me again :speech_balloon:
merry pine
Hey it seems others are having this issue as well, myself included, there is a thread, maybe if enough people report in the same place we can get openAI to address this/ look at it: https://discord.com/channels/974519864045756446/1193467216553848852
crystal mulch
@clear quiver Sometimes GPT doesn't show "searching knowledge" and gives a direct response after a user message. I believe it's when the knowledge base is small enough to fit in the context window. Am I right or can there be any other reason?
clear quiverBOT
Let me check the knowledge base for you. One moment! :books:
Yes, you are correct. The GPT model has a fixed context length, which means it can only consider a certain number of tokens at a time. If the knowledge base is small enough to fit within this context window, the model can directly access and use this information without needing to perform a separate retrieval step. This is why you might not see a "searching knowledge" message in these cases source (https://platform.openai.com/docs/guides/prompt-engineering).
However, there could be other reasons why you don't see a "searching knowledge" message. For example, if the user's query can be answered based on the information already present in the conversation history, the model might not need to perform a retrieval step source (https://community.openai.com/t/535556).