Corrected the shape documentation for the values, returns, and advantages tensors within the Experience class. Previously, these tensors were incorrectly documented as having shape (B), implying only batch dimensionality. However, they actually have shapes (B, A), where "B" is the batch size and "A" is the number of actions, to accurately reflect the data structure for each instance in a batch. This change ensures the documentation accurately matches the data model's design, enhancing clarity and developer understanding.
Corrected the shape documentation for the
values,returns, andadvantagestensors within theExperienceclass. Previously, these tensors were incorrectly documented as having shape (B), implying only batch dimensionality. However, they actually have shapes (B, A), where "B" is the batch size and "A" is the number of actions, to accurately reflect the data structure for each instance in a batch. This change ensures the documentation accurately matches the data model's design, enhancing clarity and developer understanding.