这种是运行成功了么，楼主

[Songchuanlang]

(base) root@autodl-container-b2e2118052-826a86c4:~/retinanet-pytorch-master# python train.py
Load weights model_data/retinanet_resnet50.pth.

Successful Load Key: ['backbone_net.model.conv1.weight', 'backbone_net.model.bn1.weight', 'backbone_net.model.bn1.bias', 'backbone_net.model.bn1.running_mean', 'backbone_net.model.bn1.running_var', 'backbone_net.model.bn1.num_batches_tracked', 'backbone_net.model.layer1.0.conv1.weight', 'backbone_net.model.layer1.0.bn1.weight', 'backbone_net.model.layer1.0.bn1.bias', 'backbone_net.model.layer1.0.bn1.running_mean', 'backbone_net.model.layer1.0.bn1.running_var', 'backbone_net.model.layer1.0.bn1.num_batches_tracked', ' ……
Successful Load Key Num: 354

Fail To Load Key: [] ……
Fail To Load Key num: 0

但是有一次我运行之后，迭代五次就爆了错误
Get map.
0%| | 0/4952 [00:03<?, ?it/s]
Traceback (most recent call last):
File "train.py", line 515, in
epoch_step, epoch_step_val, gen, gen_val, UnFreeze_Epoch, Cuda, fp16, scaler, save_period, save_dir, local_rank)
File "/root/retinanet-pytorch-master/utils/utils_fit.py", line 105, in fit_one_epoch
eval_callback.on_epoch_end(epoch + 1, model_train)
File "/root/retinanet-pytorch-master/utils/callbacks.py", line 196, in on_epoch_end
self.get_map_txt(image_id, image, self.class_names, self.map_out_path)
File "/root/retinanet-pytorch-master/utils/callbacks.py", line 143, in get_map_txt
results = non_max_suppression(torch.cat([outputs, classification], axis=-1), self.input_shape,
TypeError: cat() got an unexpected keyword argument 'axis'