Reproducing single image on LM-O

[taeyeopl]

I'm interested in reproducing the single image LM-O and followed the examples. I wonder do I have to download all MegaPose-GSO datasets (e.g. key_to_shard.json, shard-000000.tar) since I got these errors while running the provided examples. Do I miss something during installation?

python test.py test_dataset_name=lmoWonder3d run_id=$NAME_RUN

INFO:src.dataloader.test:Split: test for !
Traceback (most recent call last):
  File "/home/miruware/anaconda3/envs/gigapose/lib/python3.9/site-packages/hydra/_internal/instantiate/_instantiate2.py", line 92, in _call_target
    return _target_(*args, **kwargs)
  File "/data/robot/workspace/gigapose/src/dataloader/test.py", line 70, in __init__
    web_dataset = WebSceneDataset(webdataset_dir / split, depth_scale=depth_scale)
  File "/data/robot/workspace/gigapose/src/custom_megapose/web_scene_dataset.py", line 131, in __init__
    frame_index = self.load_frame_index()
  File "/data/robot/workspace/gigapose/src/custom_megapose/web_scene_dataset.py", line 144, in load_frame_index
    key_to_shard = inout.load_json(root_dir / "key_to_shard.json")
  File "/data/robot/workspace/gigapose/bop_toolkit/bop_toolkit_lib/inout.py", line 84, in load_json
    f = open(path, "r")
FileNotFoundError: [Errno 2] No such file or directory: '/data/robot/workspace/gigapose/datasets/test/key_to_shard.json'

gigapose/src/dataloader/test.py

Lines 70 to 73 in 17fcf97

	# load the dataset
	webdataset_dir = self.root_dir / self.dataset_name
	web_dataset = WebSceneDataset(webdataset_dir / split, depth_scale=depth_scale)
	self.web_dataloader = IterableWebSceneDataset(web_dataset, set_length=True)

[taeyeopl]

I partly solved the problem reference by #15 and #26.

The provided scripts (python -m src.scripts.download_cnos_bop23) look like only provide GT 3D model CNOS detection results.

Could you share the results of the single image (Wonder 3D Mesh) CNOS detection on the LM-O dataset?