Faster R-CNN运行报错：ValueError: ‘a‘ cannot be empty unless no samples are taken

Traceback (most recent call last):
  File "/root/autodl-tmp/Faster-rcnn/train.py", line 451, in <module>
    fit_one_epoch(model, train_util, loss_history, eval_callback, optimizer, epoch, epoch_step, epoch_step_val, gen, gen_val, UnFreeze_Epoch, Cuda, fp16, scaler, save_period, save_dir)
  File "/root/autodl-tmp/Faster-rcnn/utils/utils_fit.py", line 27, in fit_one_epoch
    rpn_loc, rpn_cls, roi_loc, roi_cls, total = train_util.train_step(images, boxes, labels, 1, fp16, scaler)
  File "/root/autodl-tmp/Faster-rcnn/nets/frcnn_training.py", line 321, in train_step
    losses = self.forward(imgs, bboxes, labels, scale)
  File "/root/autodl-tmp/Faster-rcnn/nets/frcnn_training.py", line 248, in forward
    rpn_locs, rpn_scores, rois, roi_indices, anchor = self.model_train(x = [base_feature, img_size], scale = scale, mode = 'rpn')
  File "/root/miniconda3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1501, in _call_impl
    return forward_call(*args, **kwarg