training on replica dataset error

[Mass17]

I run the training with replica dataset, which I have downloaded from onedrive (mentioned in the README.md)

**max_steps=100000
output_dir="./outputs/splat360_log_depth_near0.1-100k/"
CUDA_VISIBLE_DEVICES=0,1 python -m src.main \
     +experiment=replica data_loader.train.batch_size=1 \
     model.encoder.shim_patch_size=8 \
     model.encoder.downscale_factor=8 \
     trainer.max_steps=$max_steps \
     model.encoder.depth_sampling_type="log_depth" \
     output_dir=$output_dir \
     dataset.near=0.1**

But I get the following error,

File "/home/mdul/anaconda3/envs/splat360/lib/python3.10/site-packages/jaxtyping/_decorator.py", line 470, in wrapped_fn_impl
   raise TypeCheckError(msg) from e
jaxtyping.TypeCheckError: Type-check error whilst checking the parameters of src.visualization.layout.vcat.
The problem arose whilst typechecking parameter 'images'.
Actual value: (f32[3,512,1024](torch), f32[3,512,1024](torch))
Expected type: typing.Iterable[Float[Tensor, 'channel _ _']].
----------------------
Called with parameters: {
 'images': (f32[3,512,1024](torch), f32[3,512,1024](torch)),
 'align': 'start',
 'gap': 8,
 'gap_color': 1
}
Parameter annotations: (*images: Iterable[Float[Tensor, 'channel _ _']], align: Literal['start', 'center', 'end', 'left', 'right'] = 'start', gap: int = 8, gap_color: Union[int, float, Iterable[int], Iterable[float], Float[Tensor, '#channel'], Float[Tensor, '']] = 1) -> Any.

I asked from ChatGPT, but could not solve it unfortunately.
FYI: replica_dataset/test <--- only "test" folder we have if we downoad the replica dataset from onedrive.
but replica_dataset_pt has both "train" and "test" folders.

[thucz]

As a matter of fact, we didn't use Replica for training. We use HM3D for training and use HM3D & Replica for evaluation.