Remix.run Logo
stingrae 7 hours ago

the model is a set of weights, you can take a snapshot and test it. Reinforcement learning itself is largely testing and tuning.