Comment on GroqchatparentComments−tome2ynomel gave a good answer in a different threadThis is not about the model, it’s about the relative speed improvement from the hardware, with this model as a demo.To compare apples to apples look at the tokens per second of other systems running Llama 2 70B 4096. We're by far the fastest!https://news.ycombinator.com/item?id=38742466
Comments
nomel gave a good answer in a different thread
To compare apples to apples look at the tokens per second of other systems running Llama 2 70B 4096. We're by far the fastest!
https://news.ycombinator.com/item?id=38742466