Run Google Gemma 2B and 7B on your computer for free (CPU and GPU)
HTML-код
- Опубликовано: 20 фев 2024
- Hello, welcome to this video. Google just released the best open-source model on the market right now called Gemma. It uses the same architecture as Gemini but it is smaller. We have Gemma 2B and Gemma 7B.
In this video, I will show you how to launch Gemma on your local computer (GPU and CPU) and use it like your local GPT.
Tech stack: Ollama, openGui, Docker, Gemma, Huggingface.
Leave a comment if you read this text :)
Blog post with more details: ulife.ai/
You can visit our apps here :
Launchpad AI here: launchpad.ulife.ai
Follow Me:
Tiktok: / ltphen
Instagram: / ltphenvonulife
Follow Ulife here: / ulife-co 
Наука
Great video, thanks
The demo is very detailed, so hopefully I'll figure out how to switch GPUs soon!
Sure. I am preparing a tutorial for that
A good tutorial to run Gemma perfectly on a regular computer!
Thanks a lot
You're gonna be famous good luck bro
Thanks a lot man
Thanks for the tutorial! ^_^ Is there a way to have Gemma access the web to read content of web pages? Thanks :)
Yes just use #before your url. Like #www.google.com
@@ltphenvonulife thanks but it doesn't seem to work. I wrote:
"please visit #google.it and search for "gatto" and print the first 3 results"
Result Gemma:7b
Explanation of the phrase "google.it gatto":
Components: ....
No navigation :( Am I missing something?
This video is perfect,I want know can gemma be use on MacOs?Can you tell me?please!
Yes. You can use the exact same way i did on windows on Mac or Linux. Just install docker before
Thats great , but more simple way we can do , I think with GPT4All , we just need to install GPT4All then download and put gemma inside the models folder.
Ohh I haven't tried that one. I thought it wouldn't be possible as GPT4all only work for some models. I will check on my side