← Back to context Comment by cubefox 4 hours ago Yep. Zero temperature is neither necessary nor sufficient for deterministic inference. 3 comments cubefox Reply cj 4 hours ago Why? tempay 4 hours ago You can seed the randomness are still having nonzero temperature.Numerical instability can introduce randomness especially on GPU like hardware unless you’re very careful about how you write your algorithms. vessenes 2 hours ago In any batch inference environment that includes experts, expert routing may vary depending on what else is in the batch. For one thing.
cj 4 hours ago Why? tempay 4 hours ago You can seed the randomness are still having nonzero temperature.Numerical instability can introduce randomness especially on GPU like hardware unless you’re very careful about how you write your algorithms. vessenes 2 hours ago In any batch inference environment that includes experts, expert routing may vary depending on what else is in the batch. For one thing.
tempay 4 hours ago You can seed the randomness are still having nonzero temperature.Numerical instability can introduce randomness especially on GPU like hardware unless you’re very careful about how you write your algorithms.
vessenes 2 hours ago In any batch inference environment that includes experts, expert routing may vary depending on what else is in the batch. For one thing.
Why?
You can seed the randomness are still having nonzero temperature.
Numerical instability can introduce randomness especially on GPU like hardware unless you’re very careful about how you write your algorithms.
In any batch inference environment that includes experts, expert routing may vary depending on what else is in the batch. For one thing.