Kevin Hu

Results 1524 comments of Kevin Hu

Could you please paste the graph, and elaberate on a real scenario to help us understand that deeply? Realy important, thanks!

Set the max token to 128000 while adding the models. ![Image](https://github.com/user-attachments/assets/cae62252-6368-4fd7-83d2-747a3a27eaae)

Time out to invoke LLM. `export LM_TIMEOUT_SECONDS=3600`

Layouts recognition and table structure recognition all use Yolov8. Nothing special on hyper-parameters. Please refer to [this](https://yolov8.org/how-to-use-fine-tune-yolov8/) for trainning.

The system prompt in RAGFlow will be filled with retrieved content, which might be long.

tenant_llm stores authorization of tenant to LLM models. The 'Team' function will be release soon.

Did not get it. Could you elabrate on that?