Invalid RLE mask representation

[ns-tm42]

I'm trying to create a custom dataset to train a detectron2 based model, and my RLE's encoding is failing to be decoded. I am using mask.py's encode function on binary masks to create my 'counts' entry in my annotations.json.

` File "/home/ubuntu/miniconda3/envs/SSIS_rebuild/lib/python3.8/site-packages/detectron2/utils/serialize.py", line 26, in call
return self._obj(*args, **kwargs)
File "/mnt/c/.rebuild_SSIS/SSIS_pipe/SSIS/adet/data/dataset_mapper.py", line 178, in call
annos = [
File "/mnt/c/.rebuild_SSIS/SSIS_pipe/SSIS/adet/data/dataset_mapper.py", line 179, in
transform_instance_annotations(
File "/mnt/c/.rebuild_SSIS/SSIS_pipe/SSIS/adet/data/detection_utils.py", line 20, in transform_instance_annotations
annotation = d2_transform_inst_anno(
File "/home/ubuntu/miniconda3/envs/SSIS_rebuild/lib/python3.8/site-packages/detectron2/data/detection_utils.py", line 314, in transform_instance_annotations
mask = mask_util.decode(segm)
File "/home/ubuntu/miniconda3/envs/SSIS_rebuild/lib/python3.8/site-packages/pycocotools/mask.py", line 91, in decode
return _mask.decode([rleObjs])[:,:,0]

File "pycocotools/_mask.pyx", line 149, in pycocotools._mask.decode
ValueError: Invalid RLE mask representation`

[achrafs758]

@ns-tm42
Can you show me how you RLE looks like?