C cerebras.ai Llama 3.1 405B now runs at 969 tokens/s on Cerebras Inference 427 points benchmarkist 1 year ago 156 comments
