This open-source model was created by Meta AI.
You can find more details here.
The model is available on the huggingface hub: https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct.
The 70B model was pretrained on 15 trillion tokens spanning 7 different languages and fine-tuned on over 25M synthetically generated examples.
It supports up to 128K token contexts.