Thanks for sharing the code,
can you please share arguments that has been used for training, specifically in the paper propose a modified version of positional encoding but I see that it is set to use None in the file.
Thanks for your help.
pos_embed=None means the model does not use the absolute position embedding (The model is set to use the relative position embedding.). Hyper-parameters are provided in the training file for each dataset.
Thanks for sharing the code, can you please share arguments that has been used for training, specifically in the paper propose a modified version of positional encoding but I see that it is set to use None in the file. Thanks for your help.