Great work 🇺🇸
The power of open-source. A lot of people here throwing their opinion on what, when, and how you should choose to spend your resources. Appreciative of perplexity for releasing, I will be testing this week and hopefully start shifting some over where I use reasoning models.
The power of open-source. A lot of people here throwing their opinion on what, when, and how you should choose to spend your resources. Appreciative of perplexity for releasing, I will be testing this week and hopefully start shifting some over where I use reasoning models.
Educate yourself more before posting. What a shame! It seems that you guys, who are interested in politics, haven't done any research on AI before.
I want to know if they touched any of the unaligned bits or if it was a light training. If they only removed CCP bias then good.. if they add standard political correctness then that's bad. Time will tell when the model becomes available to use.
@Yuanze If by educate yourself you mean appeal to a higher authority on every comment because the actual content of your reply is baseless opinion, I am good. Almost every comment has included some form of response to an assumption you yourself made about the other person. You have zero ground to stand on when it comes to being taken seriously. You can't hide the ignorance and logical fallacies behind credentials.
@huntsyea I appreciate your honesty. It's important to note that LLMs can be fine-tuned using selectively chosen data for political purposes. We all hope that the focus of the HuggingFace community will remain on technical innovation rather than political manipulation. Especially when companies haven't disclosed/released the details of their so-called collected datasets, we certainly don't want open-source models to be repurposed as tools for political agendas.
As researchers, we do hope that both the papers and models remain driven only by innovation.
I'm not sure if you've visited China before, but experiencing it firsthand might provide a clearer and more objective perspective.
You are conflating "you" with "we".
"We certainly don't want open-source models to be repurposed as tools for political agendas." - This reads I want a select group to control when, how, and for what a model are used. That is the antithesis of open-source at a fundamental level.
"We certainly don't want open-source models to be repurposed as tools for political agendas." - That statement refers to some politicians and certain segments of Western media, not to you. I meant to encourage careful and responsible use, while keeping the open-source spirit intact.
America will win BIG TIME! President Trump no longer needs tens of billions to train any models. He just needs to dump a billion dollars into media and CIA disinformation campaigns, then feed this "unbiased" content into China's open-source AI models – and voilà, he'll instantly clinch victory in this race! 🏆🇺🇸
💪💪且听鹰哮