xiaochenchen

Results 10 comments of xiaochenchen

[BUG] 搜索返回问题

请问下解决了吗

QUESTION: `No available slot found for the model` when launching multiple models

> Hi，感谢对 xinference 的关注！ > > 前两个问题实际上是同一个问题。原因是为了降低 OOM 发生的概率，目前 xinference 的策略是每张卡上只能放一个模型。因此如果只有一块 GPU 时，启动第二个模型会报错 `No available slot found for the model`。 > > 你可以 terminate 当前模型，此时 GPU 空闲，就可以加载第二个模型了。 > > 另外，在一张卡上加载多个模型的策略正在开发中，到时如果显存足够，将能够在同一个...

BUG: Model not found in the model list, uid: {model_uid}

> same issue, and the rerank model 怎么解决的呢？

Q4_0 + CPU转换避坑指南

请问下你们有试过GPU版本的么，我试了怎么编译都是用CPU推理

[Question]: ERROR: Expecting value: line 1 column 1 (char 0)

> Same problem with Deepseek-chat and Deepseek-reasoner. hi,same probllem,how to solve it ?

[Bug]: When the knowledge graph is enabled, retrieval testing throw error: JSONDecodeError('Expecting value: line 1 column 1 (char 0)')

> Update: It seems the knowledge graph extraction use "Chat Model" in System Model Setting rather than the embedding model in Knowledge base. I switched "Chat Model" to gpt-4o-mini and...

[Bug]: ragflow/api/db/db_models.py", line 355, in lock raise Exception(f'failed to acquire lock {self.lock_name}')

> > Is that on Mac? set `MACOS=1` in `docker/.env`. Or, switch to `mariaDB` instead of `MySQL`. > > 遇到相同的问题，操作系统是：Ubuntu，数据库为mariaDB。 你好 请问解决了么

[Bug]: Knowledge graph retrieval error. It was found that many nodes in the knowledge graph would encounter errors when calling the large model, but the logs I saw were JSON errors.

> [@sammichenVV](https://github.com/sammichenVV) Could you provide: > > * The document you parsed > * The parse configuration of the knowledge base > * The query > [@sammichenVV](https://github.com/sammichenVV) Could you provide:...

[Question]: Unable to interpret file

> ![Image](https://github.com/user-attachments/assets/8bd5f99d-8533-4d4c-adf3-4b87aa2b1325) Hi, I'm facing the same issue. How can I resolve it?

[Question]: "Task is queued" for a long time.

> ### Describe your problem > I saw there was a solution to MacOS when faced with "Task is queued." However, I use Linux and Windows, with embedding model and...