That design was trained in part utilizing their unreleased R1 "thinking" design. Today they've R1 itself, together with a whole family of new designs obtained from that base.

There's a lot of things in the new release.

DeepSeek-R1-Zero appears to be the base design. It's over 650GB in size and, archmageriseswiki.com like many of their other releases, is under a clean MIT license. DeepSeek alert that "DeepSeek-R1-Zero encounters difficulties such as unlimited repeating, poor readability, and language mixing." ... so they likewise launched:

DeepSeek-R1-which "includes cold-start data before RL" and "attains performance similar to OpenAI-o1 throughout math, code, and thinking tasks". That one is also MIT accredited, and is a similar size.

I don't have the ability to run models larger than about 50GB (I have an M2 with 64GB of RAM), so neither of these two models are something I can quickly have fun with myself. That's where the new distilled models are available in.

To support the research study community, we have open-sourced DeepSeek-R1-Zero, DeepSeek-R1, asteroidsathome.net and six dense models distilled from DeepSeek-R1 based on Llama and Qwen.

This is a fascinating flex! They have actually designs based on Qwen 2.5 (14B, 32B, Math 1.5 B and Math 7B) and Llama 3 (Llama-3.1 8B and Llama 3.3 70B Instruct).

Weirdly those Llama models have an MIT license attached, which I'm uncertain works with the underlying Llama license. Qwen designs are Apache licensed so perhaps MIT is OK?

(I likewise simply noticed the MIT license files state "Copyright (c) 2023 DeepSeek" so they might require to pay a little bit more attention to how they copied those in.)

Licensing aside, these distilled models are fascinating monsters.

Running DeepSeek-R1-Distill-Llama-8B-GGUF

Quantized versions are already starting to reveal up. Up until now I've tried just one of those- unsloth/DeepSeek-R 1-Distill-Llama-8B-GGUF released by Unsloth AI-and addsub.wiki it's actually fun to have fun with.

I'm running it using the combination of Ollama, online-learning-initiative.org LLM and the llm-ollama plugin.

First I brought and ran the model using Ollama itself:

This downloads around 8.5 GB of model information and wiki.fablabbcn.org starts an interactive chat user interface.

Once the design has been brought, LLM can speak with it too. I choose utilizing LLM for experiments since it logs whatever to SQLite for later exploration.

I set up the plugin and ran it like this:

Each reaction begins with a ... pseudo-XML tag containing the chain of thought used to assist create the action. For "a joke about a pelican and a walrus who run a tea space together" that started like this:

Okay, so I need to come up with a joke about a pelican and a walrus running a tea space together. Hmm, that's a fascinating combination. Let me believe about how these two characters may interact in a humorous circumstance.

First, let's consider their characteristics. Pelicans are known for their long beaks and Webbed feet, frequently seen near the beach or water. Walruses have big teeth, thick fur, and they're generally discovered in colder environments, like icebergs or snowy areas. So, integrating these 2 into a tea room setting is currently a funny image.

[...]
It then believed for 20 paragraphs before outputting the joke! Here's the complete records and here's the output at the end, after the tag:

The Joke:

A pelican and a walrus choose to open a quaint little tea space together. The walrus, surgiteams.com with its big size, has a hard time to discover comfortable chairs, so it sits on the table by mishap, overturning the teapot. Meanwhile, the pelican, attempting to assist, utilizes its beak to put saucers on the table, triggering a few spills.

After a series of humorous mishaps, the walrus searches for and states with a grin, "This isn't so fishy anymore." The pelican smirks and says, "Maybe not, but we do have a lot of krill in our tea!"

Explanation:

- Setup: Pelican and Walrus run an unlikely tea space. Their sizes result in comedic chaos.

• Punchline: "This isn't so fishy anymore" is a pun on the circumstance becoming less absurd. The follow-up joke presents "krill," a play on words that connects into both marine life ("fish") and trademarketclassifieds.com tea, making it funny while fitting the style.
  
  So yeah, the joke is awful. But the procedure of getting there was such a fascinating insight into how these brand-new designs work.
  
  This was a fairly small 8B design. I'm eagerly anticipating trying the Llama 70B version, which isn't yet available in a GGUF I can run with Ollama. Given the strength of Llama 3.3 70B-currently my preferred GPT-4 class model that I've run on my own machine-I have high expectations.
  
  Update 21st January 2025: I got this quantized version of that Llama 3.3 70B R1 distilled model working like this-a 34GB download:
  
  Can it draw a pelican?
  
  I attempted my classic Generate an SVG of a pelican riding a bicycle prompt too. It did refrain from doing effectively:
  
  It aimed to me like it got the order of the elements incorrect, so I followed up with:
  
  the background ended up covering the remainder of the image
  
  It thought some more and offered me this:
  
  Similar to the earlier joke, the chain of believed in the records was far more fascinating than the end outcome.
  
  Other ways to attempt DeepSeek-R1
  
  If you want to attempt the model out without setting up anything at all you can do so utilizing chat.deepseek.com-you'll require to develop an account (indication in with Google, utilize an email address or offer a Chinese +86 contact number) and after that select the "DeepThink" alternative below the timely input box.
  
  DeepSeek use the model via their API, using an OpenAI-imitating endpoint. You can access that via LLM by dropping this into your extra-openai-models. yaml setup file:
  
  Then run llm keys set deepseek and paste in your API key, then utilize llm -m deepseek-reasoner 'timely' to run triggers.
  
  This won't reveal you the reasoning tokens, sadly. Those are provided by the API (example here) however LLM doesn't yet have a way to display them.