Kevin Hu comments

Results 1517 comments of


                                            Kevin Hu

I have completed the deployment of all source code and it has been successfully running, but I am still encountering a "gateway timeout" issue. Please provide guidance to resolve this problem.

I don't think ragflow-server is up. ![Image](https://github.com/user-attachments/assets/3e0079de-5847-4217-965d-5e81e168970d)

[Question]: Deepdoc pdf parsing consumes too much memory

> Can i run deepdoc on GPU? Will it be faster? For open-source Deepdoc has nothing to do with GPU.

Fix data_sync startup crash by properly invoking async main

Appreciations! Please remove the unnecessary code format alterations.

[Question]: I created a chat application using ragflow Chat and selected the knowledge base I wanted. However, when searching, I couldn't retrieve the content in the knowledge base and couldn't use it. What could be the reason for this?

Could you share us with the file and the query? [[email protected]](mailto:[email protected])

Plugin hook error

RAGFlow does not use image data or image embeddings during retrieval.

[Question]: Why can't the chat find the content in the knowledge base?

You could click the little lamp to see wether it's been retrieved? If it is, it goes to LLM. ![Image](https://github.com/user-attachments/assets/47043b96-9bd1-4680-b06b-347e330410fc)

[Question]: Why can't the chat find the content in the knowledge base?

@BogenZ Did you apply `Table` as chunking method?

[Question]: Chat - list index out of range error

What about checking embedding and LLM configuration again? I doubt either one of them is faild to invoke.

[Question]: Tool call error, please correct the input parameter of response format and call it again.

What about switching to another LLM? It's highly related to LLM NLU capability.

[Bug]: Multi-turn optimization config is invalid and can not add reranker type from lm stuido

It can not connect rerank model deployed by LmStudio for now. The duplicated answer primarily caused by the LLM if the message history sent to LLM is correct.