Closed naszka closed 2 years ago
The same prefix was not added in DataPipeline.load_data in minerl so the pipeline is looking for files without 'train'
It is easier for now to just specify the path with "train" when laoding.
data = iglu.data.make(data_dir='~/iglu_data/train')
this can break the consistency between the downloader and dataloader. I fixed that in #13.
The same prefix was not added in DataPipeline.load_data in minerl so the pipeline is looking for files without 'train'
It is easier for now to just specify the path with "train" when laoding.
data = iglu.data.make(data_dir='~/iglu_data/train')