LlamaIndex
Log in

Find answers from the community

Updated 6 months ago

I have a problem when loading more than

At a glance

The community member is experiencing an issue when loading more than one PDF document, where the program works with a single PDF but stops working with two files. The comments suggest that the issue is related to a timeout, and the community member is using the Mistral for Ollama LLM. The recommended solutions are to increase the request timeout and to consider using a GPU-based model for better performance when dealing with large data inputs.

SSebitas
·
I have a problem when loading more than one PDF document, I don't know if my implementation is correct since the times I have tried to run the program it gives me the same error, if I use a single PDF document it works if I implement this way with 2 files stops working
Attachments
image.png
image.png
L
S
9 comments
LLogan M
where is the timeout happening? When querying? loading data? Something else?
SSebitas
It happen in the while, the other code runs well
Attachment
image.png
LLogan M
ah I think that makes sense. What LLM are you using? You likely just have to increase the timeout
SSebitas
Im using Mistral for Ollama
Attachment
image.png
SSebitas
What i can do to improve the performance, increasing the request timeout is the only way?
LLogan M
Ollama is already running the model as fast as it can from my understanding

Just need to increase request timeout so that it doesn't timeout
LLogan M
Ollama is cool, but it's not really meant for large data use cases imo
LLogan M
Running models on an actual gpu is really the only way to get reasonable speeds when the inputs to the llm are large
SSebitas
Ok i understand thank you very much
Add a reply
Sign up and join the conversation on Discord