HF RL Explorer

Task: Implement TensorBoard's Exponential Moving Average Smoothing Algorithm

Task: Implement TensorBoard's Exponential Moving Average Smoothing Algorithm: a task in Terminal-Lego-15k (Harbor dataset). TensorBoard uses an exponential moving average (EMA) algorithm to smooth scalar graphs. The smoothing parameter (weight) ranges from 0 to 1, where 0 means no smoothing and 1…

The task

TensorBoard uses an exponential moving average (EMA) algorithm to smooth scalar graphs. The smoothing parameter (weight) ranges from 0 to 1, where 0 means no smoothing and 1 means maximum smoothing.

Part of PrimeIntellect/Terminal-Lego-15k.