Open SatireY opened 6 months ago
Thank you for your interest in our work. The training data we used in the paper is of the same kind as that of the 'taco_play' data in the Open-X dataset. You can download the taco_play data as described there. Then you can use the generator function generator_fun_openx_nooverlap to create the dataset by:
ds_train = Dataset.from_generator(generator_fun_openx_nooverlap, gen_kwargs={ 'data_path': "/path/to/taco/download/dir", "traj_len": 10})
Please let me know if you encounter further problems :)
Thank you for your work, I want to try to run your code, but I found that you did not provide the data set you mentioned in the paper, how should I run your code? Where should I download the data?