RNN/3D model does not work with tubplot

[mgagvani]

Example command:

donkey tubplot --tub=data/ --model=models/pilot_24-06-10_4.h5 --type=rnn --limit=30000

Results in error:

Traceback (most recent call last):
  File "/home/mgagvani/miniconda3/envs/donkey/bin/donkey", line 8, in <module>
    sys.exit(execute_from_command_line())
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/mnt/f/Projects/TensorKart-master/donkeycar/donkeycar/management/base.py", line 625, in execute_from_command_line
    c.run(args[2:])
  File "/mnt/f/Projects/TensorKart-master/donkeycar/donkeycar/management/base.py", line 522, in run
    self.plot_predictions(cfg, args.tub, args.model, args.limit,
  File "/mnt/f/Projects/TensorKart-master/donkeycar/donkeycar/management/base.py", line 474, in plot_predictions   
    user_angle = tub_record.underlying['user/angle']
                 ^^^^^^^^^^^^^^^^^^^^^
AttributeError: 'list' object has no attribute 'underlying'

Am using donkey v5.2.dev2, developer install on WSL, latest commit. This can be reproduced with both the 3D CNN and RNN for arbitrary length values.

Another minor thing I didn't want to file a whole new issue for, but donkey models shows the length value for RNN values but not for 3D CNN:

4       pilot_24-06-10_4           KerasLSTM-L:6          rnn  data/ 2024-06-10 19:28:59     None    None
5       pilot_24-06-10_5             Keras3D_CNN           3d  data/ 2024-06-10 21:14:06     None    None

[mgagvani]

Update: same thing happens with the memory model.