"More than one task name found for task id %d: error

[paul-reiners]

What would cause this error when doing inference?

Traceback (most recent call last):
  File "/home/support/public/torch_cudnn8.2/bin/nnUNet_predict", line 8, in <module>
    sys.exit(main())
  File "/home/support/public/torch_cudnn8.2/lib/python3.9/site-packages/nnunet/inference/predict_simple.py", line 151, in main
    task_name = convert_id_to_task_name(task_id)
  File "/home/support/public/torch_cudnn8.2/lib/python3.9/site-packages/nnunet/utilities/task_name_id_conversion.py", line 47, in convert_id_to_task_name
    raise RuntimeError("More than one task name found for task id %d. Please correct that. (I looked in the "
RuntimeError: More than one task name found for task id 512. Please correct that. (I looked in the following folders:
/home/feczk001/shared/data/nnUNet/nnUNet_raw_data_base/nnUNet_raw_data
/home/feczk001/shared/data/nnUNet/nnUNet_raw_data_base/nnUNet_preprocessed
/home/feczk001/shared/data/nnUNet/nnUNet_raw_data_base/nnUNet_cropped_data

Here are my environment variables:

export nnUNet_raw_data_base="/home/feczk001/shared/data/nnUNet/nnUNet_raw_data_base"
export nnUNet_preprocessed="/home/feczk001/shared/data/nnUNet/nnUNet_raw_data_base/nnUNet_preprocessed"
export RESULTS_FOLDER="/home/feczk001/shared/data/nnUNet/nnUNet_raw_data_base/nnUNet_trained_models"

I've checked and there is exactly one model with the number 512:

/home/feczk001/shared/data/nnUNet/nnUNet_raw_data_base/nnUNet_trained_models/nnUNet/3d_fullres/Task512_BCP_ABCD_Neonates_SynthSegDownsample/nnUNetTrainerV2__nnUNetPlansv2.1

Here is my call to inference:

nnUNet_predict -i ./input/input2 -o ./output -t 512 -m 3d_fullres

What am I doing wrong here? Now I did upload the 512 nnUNet_preprocessed and nnUNet_trained_models folders to a different server because I was running out of disk storage. Would that cause this problem?

[paul-reiners]

The problem was that there were two 512 tasks. One had a suffix after the 512.