[GH-ISSUE #3297] Regarding model issues after Docker deployment #2121

Closed
opened 2026-02-22 18:28:13 -05:00 by yindo · 1 comment
Owner

Originally created by @zhangzhen1163 on GitHub (Feb 20, 2025).
Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/3297

Consult me,I deployed Docker on both CentOS and Windows systems, and chose the DeepSeek-r1:7b model. However, his answer was Baidu AI Assistant. Then, I installed the exe client on this Windows machine with the same configuration, and he answered DeepSeek. What is the reason for this。

Image

Image

Originally created by @zhangzhen1163 on GitHub (Feb 20, 2025). Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/3297 Consult me,I deployed Docker on both CentOS and Windows systems, and chose the DeepSeek-r1:7b model. However, his answer was Baidu AI Assistant. Then, I installed the exe client on this Windows machine with the same configuration, and he answered DeepSeek. What is the reason for this。 ![Image](https://github.com/user-attachments/assets/9ad2bef7-866e-44e6-83a2-9ca1365168e9) ![Image](https://github.com/user-attachments/assets/364bb6f5-eba4-4935-aaf0-e1739008bd8b)
yindo closed this issue 2026-02-22 18:28:13 -05:00
Author
Owner

@timothycarambat commented on GitHub (Feb 21, 2025):

Do not ask LLMs anything about themselves. That is not how models work and not how you should "verify" information about them. They are also non-deterministic so even asking the same query can give two different answers.

Not to mention you are on the 7B distillation of Deepseek - which is highly prone to hallucinations since it is quite small.
https://docs.anythingllm.com/llm-not-using-my-docs#llms-are-not-omnipotent

@timothycarambat commented on GitHub (Feb 21, 2025): Do not ask LLMs anything about themselves. That is not how models work and not how you should "verify" information about them. They are also non-deterministic so even asking the same query can give two different answers. Not to mention you are on the 7B distillation of Deepseek - which is highly prone to hallucinations since it is quite small. https://docs.anythingllm.com/llm-not-using-my-docs#llms-are-not-omnipotent
yindo changed title from Regarding model issues after Docker deployment to [GH-ISSUE #3297] Regarding model issues after Docker deployment 2026-06-05 14:44:40 -04:00
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Mintplex-Labs/anything-llm#2121