Wow, bold choice to ban the import of technology and knowledge. Usually governments are worried about export, so it doesn’t fall into the wrong hands.
Btw, how is the Nvidia stock price doing?
Right? Like, seriously, we all know somebody is just butthurt because their stock options tanked.
Oh, wait, I’m sorry! That was very unpatriotic of me, wasn’t it? I mean, we all know that winning an election guarantees being heavily rewarded with insider trading, right? It’s not like they’re there to represent constituents or anything; I mean, doesn’t everyone know we’re a republic, not a democracy?!
Sigh…
To be fair, this is common practice. Countries do this all the time to protect their economies. Mostly known in the West is China which banned many US services.
Of course, security of the data of the citizens is also a factor. You don’t want foreign countries use this data to interfere in any way.
Honestly, I don’t think this is common practice in non-oppressive countries. I mean sure, this happens in North Korea, Iran, China… But I’m relatively free to consume what I want with a few minor exceptions. For example we don’t import food that isn’t food-safe by our standards. Regardless if it’s common practice to eat it in other places. Also food may not be able to enter the country due to laws on animal cruelty. Similar things apply to electronic devices that aren’t up to code. And some select few things are banned altogether and you can’t have them and neither can someone import them. Other than that, regulations aren’t super strict. I can use all American social media platforms despite them stealing my personal data and violating European privacy laws regularly, can use Russian or Chinese websites… I think I live in a free country.
Helping domestic economy is done with tariffs / import tax. And not by banning things and putting people in jail.
And mind that this isn’t about the service that collects your data and gives it to the Chinese government. This is about downloading the model file and using it all by yourself. So no data gets transferred to a foreign country. And it’s not because people could get harmed or anything. This is just because the vice president doesn’t want it personally. Like in some dictatorship. Otherwise they would have banned transferring data into foreign countries, if that’s what it’s about. But they didn’t do that, because it’s not about protecting the people.
Or did I miss something and there are other examples for limitations on import?
now i gotta download something i don’t even wanna download.
Yup. Downloaded 7b, 32b, and 70b varieties this afternoon. Entirely out of spite.
Since those smaller models are technically fine-tunes of Meta/Facebook’s LLAMA, using Deepseek’s outputs, I wonder if they would be covered by the bill at all.
Fascist regime and power/police abuse has started.
P.S.: It seems like the US is becoming similar to Russia, kleptocratic country and organised crime in government.
For Base Model
git lfs install git clone https://huggingface.co/deepseek-ai/DeepSeek-V3-Base
For Chat Model
git lfs install git clone https://huggingface.co/deepseek-ai/DeepSeek-V3
this is deepseek-v3. deepseek-r1 is the model that got all the media hype: https://huggingface.co/deepseek-ai/DeepSeek-R1
Base models are general purpose language models, mainly useful for AI researchers and people who want to build on top of them.
Instruct or chat models are chatbots. They are made by fine-tuning base models.
The V3 models linked by OP are Deepseek’s non-reasoning models, similar to Claude or ChatGPT4o. These are the “normal” chatbots that reply with whatever comes to their mind. Deepseek also has a reasoning model, R1. Such models take time to “think” before supplying their final answer; they tend to give better performance for stuff like math problems, at the cost of being slower to get the answer.
It should be mentioned that you probably won’t be able to run these models yourself unless you have a data center style rig with 4-5 GPUs. The Deepseek V3 and R1 models are chonky beasts. There are smaller “distilled” forms of R1 that are possible to run locally, though.
I heard people saying they could run the r1 32B model on moderate gaming hardware albeit slowly
https://www.deepseekv3.com/en/download
I was assuming one was pre-trained and one wasn’t but don’t think that’s correct and don’t care enough to investigate further.
Is that website legit? I’ve only ever seen https://www.deepseek.com/
And I would personally recommend downloading from HuggingFace or Ollama
r1 is lightweight and optimized for local environments on a home PC. It’s supposed to be pretty good at programming and logic and kinda awkward at conversation.
v3 is powerful and meant to run on cloud servers. It’s supposed to make for some pretty convincing conversations.
R1 isn’t really runnable with a home rig. You might be able to run a distilled version of the model though!
