val 问题

[wang7393]

验证模型的时候随机性的报如下错误，有遇到过没 ting Model ---- Traceback (most recent call last): File "main.py", line 118, in val(model, val_dataloader, epoch, opt, val_logger, visualizer) File "/opt/conda/lib/python3.6/site-packages/torch/autograd/grad_mode.py", line 43, in decorate_no_grad return func(*args, *kwargs) File "/wangxiaobo_work/10-torch/yolo_nano/val.py", line 50, in val true_positives, pred_scores, pred_labels = [np.concatenate(x, 0) for x in list(zip(sample_matrics))] ValueError: not enough values to unpack (expected 3, got 0)

[pzxdd]

数据集格式不对吧

[Xerus-git]

我也遇到一樣的問題，請問大老們有解嘛?

[Interesting6]

应该是非极大值抑制里面未考虑无物体的情况