[Feature Request]: The Qwen3-Reranker-8B model deployed by vllm cannot be added.
Self Checks
- [x] I have searched for existing issues search for existing issues, including closed ones.
- [x] I confirm that I am using English to submit this report (Language Policy).
- [x] Non-english title submitions will be closed directly ( 非英文标题的提交将会被直接关闭 ) (Language Policy).
- [x] Please do not modify this template :) and fill in all the required fields.
Is your feature request related to a problem?
提示 : 102
Fail to access model(Qwen/Qwen3-Reranker-8B).'NoneType' object is not iterable
The vllm deployment has been completed. The above is the error message for the connection. The current version is 0.19.
Describe the feature you'd like
Support Qwen3-Reranker-8B as soon as possible
Describe implementation you've considered
No response
Documentation, adoption, use case
Additional information
No response
@kou1584423647 Noted.
RAGFlow does not care what kind of reranker model is used when reranker is deployed behind vllm. You can just choose cohere API which provided by vllm such that Qwen3-reranker can be accessed through RAGFlow
Could you please paste the back-end error log stack here?
2025-06-11 15:52:03,586 INFO 20 192.168.0.6 - - [11/Jun/2025 15:52:03] "POST /v1/llm/add_llm HTTP/1.1" 200 - 2025-06-11 15:52:07,310 INFO 20 HTTP Request: POST http://192.168.23.247:2026/rerank "HTTP/1.1 200 OK" 2025-06-11 15:52:07,311 ERROR 20 Fail to access model(Qwen/Qwen3-Reranker-0.6B).'NoneType' object is not iterable NoneType: None 2025-06-11 15:52:07,311 INFO 20 192.168.0.6 - - [11/Jun/2025 15:52:07] "POST /v1/llm/add_llm HTTP/1.1" 200 - 2025-06-11 15:52:11,457 INFO 20 HTTP Request: POST http://192.168.23.247:2026/rerank "HTTP/1.1 200 OK" 2025-06-11 15:52:11,457 ERROR 20 Fail to access model(Qwen/Qwen3-Reranker-0.6B).'NoneType' object is not iterable
How can I add a large model I built with vllm in ragflow
Give it a try.
Give it a try.
I tried for OpenAI compatible API and it shows