Article: https://proton.me/blog/deepseek
Calls it “Deepsneak”, failing to make it clear that the reason people love Deepseek is that you can download and it run it securely on any of your own private devices or servers - unlike most of the competing SOTA AIs.
I can’t speak for Proton, but the last couple weeks are showing some very clear biases coming out.
You must log in or register to comment.
How is this Open Source? The official repository https://github.com/deepseek-ai/DeepSeek-R1 contains images only, a PDF file, and links to download the model. I don’t see any code. What exactly is Open Source here? And if so, where to get the source code?
In deep learning generally open source doesn’t include actual training or inference code. Rather it means they publish the model weights and parameters (necessary to run it locally/on your own hardware) and publish academic papers explaining how the model was trained. I’m sure Stallman disagrees but from the standpoint of deep learning research DeepSeek definitely qualifies as an “open source model”
So “Open Source” to AI is just releasing a .psd file used to export a jpeg, and you need some other proprietary software like Photoshop in order to use it.
What other proprietary software is necessary to use model weights?
DeepSeek is open source, meaning you can modify code(new window) on your own app to create an independent — and more secure — version. This has led some to hope that a more privacy-friendly version of DeepSeek could be developed. However, using DeepSeek in its current form — as it exists today, hosted in China — comes with serious risks for anyone concerned about their most sensitive, private information.
Any model trained or operated on DeepSeek’s servers is still subject to Chinese data laws, meaning that the Chinese government can demand access at any time.
What??? Whoever wrote this sounds like he has 0 understanding of how it works. There is no “more privacy-friendly version” that could be developed, the models are already out and you can run the entire model 100% locally. That’s as privacy-friendly as it gets.
“Any model trained or operated on DeepSeek’s servers are still subject to Chinese data laws”
Operated, yes. Trained, no. The model is MIT licensed, China has nothing on you when you run it yourself. I expect better from a company whose whole business is on privacy.
To be fair, most people can’t actually self-host Deepseek, but there already are other providers offering API access to it.
There are plenty of step-by-step guides to run Deepseek locally. Hell, someone even had it running on a Raspberry Pi. It seems to be much more efficient than other current alternatives.
That’s about as openly available to self host as you can get without a 1-button installer.
Those are not deepseek R1. They are unrelated models like llama3 from Meta or Qwen from Alibaba “distilled” by deepseek.
This is a common method to smarten a smaller model from a larger one.
Ollama should have never labelled them deepseek:8B/32B. Way too many people misunderstood that.
The 1.5B/7B/8B/13B/32B/70B models are all officially DeepSeek R1 models, that is what DeepSeek themselves refer to those models as. It is DeepSeek themselves who produced those models and released them to the public and gave them their names. And their names are correct, it is just factually false to say they are not DeepSeek R1 models. They are.
The “R1” in the name means “reasoning version one” because it does not just spit out an answer but reasons through it with an internal monologue. For example, here is a simple query I asked DeepSeek R1 13B:
Me: can all the planets in the solar system fit between the earth and the moon?
DeepSeek: Yes, all eight planets could theoretically be lined up along the line connecting Earth and the Moon without overlapping. The combined length of their diameters (approximately 379,011 km) is slightly less than the average Earth-Moon distance (about 384,400 km), allowing them to fit if placed consecutively with no required spacing.
However, on top of its answer, I can expand an option to see its internal monologue it went through before generating the answer, which you can find the internal monologue here because it’s too long to paste.
What makes these consumer-oriented models different is that that rather than being trained on raw data, they are trained on synthetic data from pre-existing models. That’s what the “Qwen” or “Llama” parts mean in the name. The 7B model is trained on synthetic data produced by Qwen, so it is effectively a compressed version of Qen. However, neither Qwen nor Llama can “reason,” they do not have an internal monologue.
This is why it is just incorrect to claim that something like DeepSeek R1 7B Qwen Distill has no relevance to DeepSeek R1 but is just a Qwen model. If it’s supposedly a Qwen model, why is it that it can do something that Qwen cannot do but only DeepSeek R1 can? It’s because, again, it is a DeepSeek R1 model, they add the R1 reasoning to it during the distillation process as part of its training. (I think they use the original R1 to produce the data related to the internal monologue which it is learns to copy.)
What makes these consumer-oriented models different is that that rather than being trained on raw data, they are trained on synthetic data from pre-existing models. That’s what the “Qwen” or “Llama” parts mean in the name. The 7B model is trained on synthetic data produced by Qwen, so it is effectively a compressed version of Qen. However, neither Qwen nor Llama can “reason,” they do not have an internal monologue.
You got that backwards. They’re other models - qwen or llama - fine-tuned on synthetic data generated by Deepseek-R1. Specifically, reasoning data, so that they can learn some of its reasoning ability.
But the base model - and so the base capability there - is that of the corresponding qwen or llama model. Calling them “Deepseek-R1-something” doesn’t change what they fundamentally are, it’s just marketing.
