Eval.py script error

[BY-Chai]

The evaluation script has an indexing error, when executing it after training.

The evaluation is executed using the following series of commands:

cd ./scripts
python hdf5.py --dataset pi_tcn --history_length 20 --unroll_length 10
python train.py --batch_size 1024 --model_type tcn --history_length 20 --unroll_length 10 --predictor_type velocity
python train.py --batch_size 1024 --model_type tcn --history_length 20 --unroll_length 10 --predictor_type attitude
python.eval.py

The error message after eval.py is

Parsing arguments ...
Seed set to 10
/home/user/Documents/longhorizondynamicsrepo/src/FW-DYNAMICS_LEARNING/resources/experiments/20241225-233641_1/
Testing Dynamics model: /home/lpenguin/Documents/UTAT/Rao2024/src/FW-DYNAMICS_LEARNING/resources/experiments/20241225-233641_1/checkpoints/model-epoch=81-val_loss=0.00.pth
Loading arguments ...
Generating test data ...
Traceback (most recent call last):
  File "eval.py", line 114, in <module>
    main(args, hdf5_files, model_path)
  File "eval.py", line 27, in main
    test_dataloaders = [load_dataset("test", data_path, hdf5_file, args, 0, False)[1] for hdf5_file in hdf5_files] # load_dataset(mode, data_path, hdf5_file, args, num_workers, pin_memory)
  File "eval.py", line 27, in <listcomp>
    test_dataloaders = [load_dataset("test", data_path, hdf5_file, args, 0, False)[1] for hdf5_file in hdf5_files] # load_dataset(mode, data_path, hdf5_file, args, num_workers, pin_memory)
  File "/home/lpenguin/Documents/UTAT/Rao2024/src/FW-DYNAMICS_LEARNING/scripts/dynamics_learning/data.py", line 44, in load_dataset
    dataset = DynamicsDataset(data_path, hdf5_file, args)
  File "/home/lpenguin/Documents/UTAT/Rao2024/src/FW-DYNAMICS_LEARNING/scripts/dynamics_learning/data.py", line 22, in __init__
    self.state_len = self.X_shape[2] - 4 # State vector doesn't contain control inputs, so decrement
IndexError: tuple index out of range

Package versions installed via requirements.txt:
numpy==1.24.3
matplotlib==3.7.5
pandas==2.0.3
scipy==1.10.1
scikit-learn==1.3.2
h5py==3.11.0
pytorch==2.3.1 (with CUDA 12.1)
lightning==2.3
Additionally, I installed the following, which was required but not listed in requirements.txt
wandb==0.19.1