seominseok0429 / Implicit-Stacked-Autoregressive-Model-for-Video-Prediction

Implicit Stacked Autoregressive Model for Video Prediction (official implementation)
84 stars 6 forks source link

How many epochs did you train in your moving-mnist experiment? #11

Open simuzisen opened 1 year ago

simuzisen commented 1 year ago

Your paper says that your model ran 10000 epochs, but the data from the Simvp model you compared is 2000 epochs

simuzisen commented 1 year ago

On your GitHub, it was stated that your 2000 epochs of mse were 17.2, while your paper only had 15.3 for 10000 epochs. However, compared to the 2000 epochs of simvp training, the number of training epochs is not consistent.