Assertion error with grayscale images

[cleherny]

I want to train a model with a dataset of grayscale images with detectron2.
I specified in the configuration file: INPUT.FORMAT: 'L', but an assertion error was raised at at the end of the training process:

Corresponding to these lines of code :

detectron2/detectron2/engine/defaults.py

Lines 280 to 295 in 0ae803b

	def __init__(self, cfg):
	self.cfg = cfg.clone() # cfg can be modified by model
	self.model = build_model(self.cfg)
	self.model.eval()
	if len(cfg.DATASETS.TEST):
	self.metadata = MetadataCatalog.get(cfg.DATASETS.TEST[0])
	checkpointer = DetectionCheckpointer(self.model)
	checkpointer.load(cfg.MODEL.WEIGHTS)
	self.aug = T.ResizeShortestEdge(
	[cfg.INPUT.MIN_SIZE_TEST, cfg.INPUT.MIN_SIZE_TEST], cfg.INPUT.MAX_SIZE_TEST
	)
	self.input_format = cfg.INPUT.FORMAT
	assert self.input_format in ["RGB", "BGR"], self.input_format

Has anyone else encountered this problem when setting the L format?
I trained a model with grayscale images by keeping INPUT.FORMAT:'RGB' and the process went through.