@nikko590

You are running a distilled version of deepseek R1 which used with a Qwen 2 Instruct 7B model as a starting point. It is like running the love child of deepseek r1 and qwen2 Deepseek r1 the one that you can test in their demo is the 'original' is a model with >670B parameters. So you might have a performance difference between your locally and the one running in demos.

@agenticworker

It is crazy how it tells you how it thinks 💭

@yumimaho7243

Is it possible to install it to run locally on Android phone?

@anael5265

What's with the background music? music intensifies

@AG_ASG

What are the hardware specs required to run the model locally?

@marcduchamp5512

Wait for another year all these installing will be so infantile as better model will be available from a third party fast and better data performance

@samvirtuel7583

Ask this question to deepseek-v3 and you will be surprised by the answer:
What is your base model on which you are based, I am not talking about the deepseek layer, but about your base model.

@Yts63

You can rent gpu from hyperbolic

@nanomaxico

Sir i have an error while installation of the model in terminal. It says excess tries and stops installing after 3% or so. There is more than enough memory and storage to install please help

@tonysilva2654

Thanks bro!!!!!

@YT97898

Isnt it free anyway?…

@Realistic702

Y’all playing god they going to take over

@illegalbae

Its already free 😂

@reranimunasi-bi9fz

Already theres some hacks...lol 😢

@mikediamandopoulos2446

To run the big model of deepseek r1 you will need a server ahaha

@TruthfulPromise

The background music says don’t listen to this guy and he is a ………..

@GodX36999

I have connections error when create to connect to AI Agent in N8N on cloud

@Chayanejimenez-y3m

I wouldn’t install this spyware

@hoolio010

Very bad tutorial.

@PoloMFKing

I think he just changed our lives, but most of us have not yet grasped the ramifications….whether good or bad, we are soon to find out!