Wake up babe, Nvidia just casually dropped a 70B model that beats GPT-4o and Claude 3.5 Sonnet.
And it's fucking open source.
Try it: https://build.nvidia.com/nvidia/llama-3_1-nemotron-70b-instruct
Or
https://huggingface.co/chat/models/nvidia/Llama-3.1-Nemotron-70B-Instruct-HF
Or
MLX (image attached) 
Yooo
๐
๐ฎ
This is the way
it doesnt count when you train on benchmarks
weird take, but ok lol
He's not wrong, though... ๐
The cursor.ai team have a good explanation of the nuance here:
Ahh I hadnโt thought about it from the coding aspect lol
Based. If only Nvidia would play nice with their open source Linux drivers, now... ๐ค
A Nvidia lanรงou uma IA pequena e de cรณdigo aberto, que bate de frente e atรฉ supera em alguns pontos o ChatGPT e o Claude.
Oh SHIT
Not better than ChatGPT, based on one question I just asked.
Wait, I can use an OpenAI API key to run this model?
Since when was that a thing?
