I’m not too informed about DeepSeek. Is it real open-source, or fake open-source?
It’s semi-open, not fully open source as what is typically thought of.
That sounds like fake open-source. Can I download the source, build it, have the thing run locally on my own machine, and use it without it having to interact with this company’s servers?
You can’t built it yourself but you can download the model and run it locally on your machine without it interacting to any server.
There is a community-driven project aimed at making a fully reproducible version called Open-R1. You can find it at https://github.com/huggingface/open-r1.
Deepseek is the company, R1 is an MIT-licensed produce, they have the Qwen models under Apache license.
You can download, modify, run locally. There are many copies online out of Deepseek’s control.
I just looked it up.
"The AI research company DeepSeek recently released its large language model (LLM) under the MIT License, providing model weights, inference code, and technical documentation. However, the company did not release its training code, sparking a heated debate about whether DeepSeek can truly be considered “open-source.”
This controversy stems from differing interpretations of what constitutes open-source in the context of large language models. While some argue that without training code, a model cannot be considered fully open-source, others highlight that DeepSeek’s approach aligns with industry norms followed by leading AI companies like Meta, Google, and Alibaba."
So fake open-source.
