5 ms·
Groq Engineer here as well; we actually built our compiler to compile pytorch, TensorFlow, and Onnx natively, so a lot of the amazing work being done by y'all i
by matanyal 3y ago
Groq Engineer here as well; we actually built our compiler to compile pytorch, TensorFlow, and Onnx natively, so a lot of the amazing work being done by y'all isn't building much of a moat. We got LLama2 working on our hardware in just a couple of days!