Dqn-per does not use importance sampling weight in training。

[xunyiljg]

Dqn_per has no important sampling weight in training, which should be a problem?

[dnddnjs]

Thanks for the report. The file/algorithm this issue references was removed or rewritten during the #122 modernization (TF/Keras → PyTorch, curated algorithm set of 9). The original code path no longer exists, so closing as no longer applicable. Feel free to reopen if you encounter a similar issue on the current PyTorch implementation.