@misc{indiciaed680a2b6bf2e,
  title = {Enabling High-Sparsity Foundational Llama Models with Efficient Pretraining and Deployment},
  author = {Abhinav Agarwalla and Abhay Gupta and Alexandre Marques and Shubhra Pandit and Michael Goin and Eldar Kurtic and Kevin Leong and Tuan Nguyen and Mahmoud Salem and Dan Alistarh and Sean Lie and Mark Kurtz},
  year = {2024},
  url = {https://arxiv.org/abs/2405.03594},
  note = {Source identifier: 2405.03594}
}