Sign In

Ollama (Llama 3.1) Install Guide - Use Llama3.1 locally and in ComfyUI for FREE

30
Ollama (Llama 3.1) Install Guide - Use Llama3.1 locally and in ComfyUI for FREE

Hello ♥

Here is a little guide to install and use Ollama locally for free.

Please download and check the documentation here:
https://ollama.com/

Step 1

Download Ollama

Step 2

Install Ollama

Step 3

Start Ollama

check if Ollama is running

Step 4

run cmd Command Prompt

start ollama with

ollama serve

Step 5

Install Model

There are diffrent Models u can install.

We will install the newest Llama Version 3.1

hit

ollama run llama3.1

now he will install the Model.

if its done, you now have installed Llama3.1 and can use it locally.

Congratz! :)

If you want to start Ollama again after closing it, repeat Step 3, 4 and 5.

To use this in Comfyui:

https://github.com/stavsap/comfyui-ollama

you can download the little workflow i create for you.

Hope this was helpful. Please tell me in the comments if you have ever used llama or which LLM you use the most ♥

Fun Fact

i really have no idea about programming. I feel like a hacker when I enter something into command prompt and something really useful happens.

Maybe you recognized yourself too 🫠

for whatever reason you want to show me appreciation, you can: ❤️Ko-Fi❤️

30

Comments

serget2's Avatar

And if you already have it in Linux (Or Linux on Windows, yeah that is a possibility) and just want to update the Ollama version
Run Ubuntu or whatever flavour you have and in the CLI , after running,
Sudo apt-get update and sudo apt-get Upgrade (updating/grading your linux and repositories so you definitely have the latest)
simply run this command:
curl -fsSL https://ollama.com/install.sh | sh

Show more
FradyPriva's Avatar

According to my experience (which is 1 day of using LLM) best of all writes Promt Gemini, it gives results better than Llama 3.1.

Also Llama itself often bugs in other languages

Der_Zerfleischer's Avatar

Genial, gleich mal ausprobiert ;-)

Thanks for the share. Has anyone used/installed locally Mistral-Nemo-12B-Base and how does it compare to Llama Version 3.1?

What about WizardCoder for coding? Anyone try it?

AIDigitalMediaAgency's Avatar

so far, i just starting with some Llama test. will share my feedback if i have some experience :)

Pinkielicious's Avatar

Yay thankies a lot for the guide 🩷🩷🩷

AIDigitalMediaAgency's Avatar

jeffthomann871's Avatar

it'd be helpful if the png you uploaded had a workflow in it.

serget2's Avatar

I already had them with Griptape, but I am not very fond of the Llama models at least not for vision they really suck at image recognition even though they are boasted to do just that, for making prompts and archiving said prompts thaty are sufficiantly intelligent, but they are good to keep costs down, Those bigger models like ChatGPT, Claude, Gemini API's are hella expensive to run 24/7
I was hoping Deepseek or Janus would be better but they aren't yet

Show more