eaky_relu_ but it isn‘t a special case. Argument types: Tensor, bool,

YOLOv5  2024-7-5 Python-3.9.19 torch-2.0.0+cu118 CUDA:2 (NVIDIA GeForce RTX 4070 Laptop GPU, 8188MiB)
YOLOv5s summary: 365 layers, 7143196 parameters, 7143121 gradients, 38.5 GFLOPs

Transferred 8/442 items from yolov5s.pt
AMP: checks failed , disabling Automatic Mixed Precision. See https://github.com/ultralytics/yolov5/issues/7908
optimizer: SGD(lr=0.01) with parameter groups 57 weight(decay=0.0), 117 weight(decay=0.0005), 96 bias
train: Scanning F:\yolov5master\Exdark\labels\train.cache... 6626 images, 1 backgrounds, 1 corrupt: 100%|██████████| 6626/6626 [00:00<?, ?it/s]
train: WARNING  F:\yolov5master\Exdark\images\train\2015_05337.png: ignoring corrupt image/label: non-normalized or out of bounds coordinates [      1.075]
val: Scanning F:\yolov5master\Exdark\labels\val.cache... 737 images, 0 backgrounds, 0 corrupt: 100%|██████████| 737/737 [00:00<?, ?it/s]

AutoAnchor: 4.79 anchors/target, 1.000 Best Possible Recall (BPR). Current anchors are a good fit to dataset 
Plotting labels to runs\train\exp7\labels.jpg... 
Image sizes 640 train, 640 val
Using 4 dataloader workers
Logging results to runs\train\exp7
Starting training for 100 epochs...

      Epoch    GPU_mem   box_loss   obj_loss   cls_loss  Instances       Size
       0/99      11.9G     0.1074    0.06137    0.06815        137        640:   0%|          | 0/415 [00:34<?, ?it/s]WARNING  TensorBoard graph visualization failure 0 INTERNAL ASSERT FAILED at "C:\\actions-runner\\_work\\pytorch\\pytorch\\builder\\windows\\pytorch\\torch\\csrc\\jit\\ir\\alias_analysis.cpp":621, please report a bug to PyTorch. We don't have an op for aten::leaky_relu_ but it isn't a special case.  Argument types: Tensor, bool, 

Candidates:
    aten::leaky_relu_(Tensor(a!) self, Scalar negative_slope=0.01) -> Tensor(a!)
       0/99      14.1G     0.1082    0.04854    0.06871         91        640:   1%|          | 5/415 [10:25<15:15:08, 133.92s/it]

GPU不允许的原因，求助！！！！！！