Hacker News Reader

I've tried out DeepSeek on deepseek.com and it refuses conversations about several topics censored in China (Tiananmen, Xi Jinping as Winnieh-the-Pooh).

Has anyone tried if this also happens when self-hosting the weights?

View on HN

I haven't tried that base model yet but I have tried with the coder model before and experienced similar things. A lot of refusals to write code if the model thought that it was unethical or could be used unethically. Like asking it to write code to download images from an image gallery website would work or not depending on what site it thought it was going to retrieve from.

I just tried the GGUF 7b model of Deepseek and it let me ask some questions about some pretty sensitive topics - Uighur Muslims, Tank man, etc.

https://huggingface.co/TheBloke/deepseek-llm-7B-chat-GGUF

When I try out the topics you suggest at the huggingface endpoint you link, the answer is either my question translated into Chinese, or no answer when I prompt the model in Chinese:

<User>: 历史上的“天安门广场的坦克人”有什么故事？ <Assistant>:

Interesting - I can't speak to the Huggingface endpoint. I downloaded the 4-bit GGUF model locally and ran it through Oobabooga with instruct-chat template - I expressed my questions in English.

Reply on news.ycombinator.com