Error "same number of tiles in inputs and target"

[toren77]

robosat$ ./rs train --model config/model-unet.toml --dataset config/dataset-buildings-dar-es-salam.toml

len(inputs)= 60859
len(inputs[0].tiles)= 60859
target.tiles[0]= (Tile(x=638131, y=543981, z=20), 'data/dar-es-salam-buldings-dataset/training/labels/20/638131/543981.png')
len(target)= 60859

len(inputs)= 7798
len(inputs[0].tiles)= 7798
target.tiles[0]= (Tile(x=638131, y=544010, z=20), 'data/dar-es-salam-buldings-dataset/validation/labels/20/638131/544010.png')
len(target)= 8453

Traceback (most recent call last):
File "/anaconda3/envs/robosat/lib/python3.6/runpy.py", line 193, in _run_module_as_main
"main", mod_spec)
File "/anaconda3/envs/robosat/lib/python3.6/runpy.py", line 85, in _run_code
exec(code, run_globals)
File "/robosat/robosat/tools/main.py", line 57, in
args.func(args)
File "/robosat/robosat/tools/train.py", line 73, in main
train_loader, val_loader = get_dataset_loaders(model, dataset)
File "/robosat/robosat/tools/train.py", line 201, in get_dataset_loaders
[os.path.join(path, "validation", "images")], os.path.join(path, "validation", "labels"), transform
File "/robosat/robosat/datasets.py", line 62, in init
assert len(self.target) == len(self.inputs[0]), "same number of tiles in inputs and target"
AssertionError: same number of tiles in inputs and target

[daniel-j-h]

The number of tiles in your slippy map directory with images has to be the same number of tiles with corresponding masks. This has to hold for both the training as well as validation dataset.

Quick check e.g. via

find dataset/training/images -type f | wc -l
find dataset/training/labels -type f | wc -l