Error on Analyze / Empty logits

[FredOMad]

Whenever I am trying to use the analyze function for my experiments (CTX with checkpoints eval on omniglot) it crashes. Loading the tfevents files on tensorboard shows them seeming to be empty too.

Traceback (most recent call last):
  File "/opt/conda/lib/python3.8/runpy.py", line 194, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/opt/conda/lib/python3.8/runpy.py", line 87, in _run_code
    exec(code, run_globals)
  File "/home/jovyan/meta-dataset/meta-dataset/meta_dataset/analyze.py", line 640, in <module>
    tf.app.run(main)
  File "/home/jovyan/venv/MD-TFDS/lib/python3.8/site-packages/tensorflow/python/platform/app.py", line 40, in run
    _run(main=main, argv=argv, flags_parser=_parse_flags_tolerate_undef)
  File "/home/jovyan/venv/MD-TFDS/lib/python3.8/site-packages/absl/app.py", line 303, in run
    _run_main(main, args)
  File "/home/jovyan/venv/MD-TFDS/lib/python3.8/site-packages/absl/app.py", line 251, in _run_main
    sys.exit(main(argv))
  File "/home/jovyan/meta-dataset/meta-dataset/meta_dataset/analyze.py", line 633, in main
    analyze_events(paths_to_event_files, FLAGS.root_dir,
  File "/home/jovyan/meta-dataset/meta-dataset/meta_dataset/analyze.py", line 549, in analyze_events
    shot_to_precision = get_shot_to_precision(shots, test_logits,
  File "/home/jovyan/meta-dataset/meta-dataset/meta_dataset/analyze.py", line 171, in get_shot_to_precision
    episode_logits = logits[episode_num]
IndexError: list index out of range

I am still getting the mean as a result at the end of the evaluation, but might this issue with analyze be related to this error at the end of the eval run?

I1219 12:05:56.362564 140662583764800 train.py:238] Starting evaluation for dataset(s) ['omniglot']...
I1219 12:05:56.362835 140662583764800 trainer.py:1534] Performing evaluation of the test split using 10 episodes...
I1219 12:07:19.650777 140662583764800 trainer.py:1553] Finished evaluation.
I1219 12:07:19.652327 140662583764800 trainer.py:1564] Accuracy on the meta-test split: 0.731521, +/- 0.082894.

2021-12-19 12:07:20.450865: W tensorflow/core/kernels/data/generator_dataset_op.cc:107] Error occurred when finalizing GeneratorDataset iterator: FAILED_PRECONDITION: Python interpreter state is not initialized. The process may be terminated.
     [[{{node PyFunc}}]]