Evaluation stuck at 0% when trying to finetune OD model

I’m trying to base a training loop using this script but I get the following issue:

50%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–Œ                                                                                  | 77/154 [00:52<00:43,  1.77it/s][INFO|trainer.py:3829] 2024-09-03 02:07:25,225 >> 
***** Running Evaluation *****
[INFO|trainer.py:3831] 2024-09-03 02:07:25,225 >>   Num examples = 305
[INFO|trainer.py:3834] 2024-09-03 02:07:25,225 >>   Batch size = 8

  0%|                                                                                                                                                                               | 0/20 [00:00<?, ?it/s]

Where after the first epoch, the evaluation loop tries to start but it stays at 0% without any movement and not even an error. The GPU usage meanwhile is at 100% while this is happening, and the memory is also being consumed.

Is there a way to add even more debug logs, or has someone come across this issue before please? Thanks.