I am reading the introduction of policy gradients by OpenAI SpinningUp. It highlights the difference between the loss function in the supervised setting and the “loss” […]
every time i bring this up people tell me that only flagships are expensive, that's not true because the definition of a midrange has changed, just […]