loss出现nan

[YCG1]

自己制作的9个类别的数据集
train: Fast image access (ping: 0.00.0 ms, read: 2391.01168.9 MB/s, size: 594.6 KB)
train: Scanning C:\Users\12818\Desktop\software\QAT.Ultralytics\coco\labels\train2017.cache... 204 images, 1 backgrounds, 0 corrupt: 100%|██████████| 204/204 [00:00<?, ?it/s]
val: Fast image access (ping: 0.00.0 ms, read: 1370.9526.7 MB/s, size: 664.7 KB)
val: Scanning C:\Users\12818\Desktop\software\QAT.Ultralytics\coco\labels\val2017.cache... 52 images, 0 backgrounds, 0 corrupt: 100%|██████████| 52/52 [00:00<?, ?it/s]
Plotting labels to runs\detect\train5\labels.jpg...
optimizer: SGD(lr=1e-05, momentum=0.937) with parameter groups 167 weight(decay=0.0), 174 weight(decay=0.0005), 173 bias(decay=0.0)
Image sizes 640 train, 640 val
Using 8 dataloader workers
Logging results to runs\detect\train5
Starting training for 6 epochs...

  Epoch    GPU_mem         lr   box_loss   cls_loss   dfl_loss  Instances       Size
    1/6      2.19G      1e-05        nan        nan        nan          9        640:   0%|          | 1/204 [00:00<01:40,  2.01it/s]

Traceback (most recent call last):
File "C:\Users\12818\Desktop\software\QAT.Ultralytics\train.py", line 10, in
results = model.train(data="mycoco.yaml", epochs=6, imgsz=640) #, device=[0, 1]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\12818\Desktop\software\QAT.Ultralytics\ultralytics\engine\model.py", line 872, in train
self.trainer.train()
File "C:\Users\12818\Desktop\software\QAT.Ultralytics\ultralytics\engine\trainer.py", line 239, in train
self._do_train(world_size)
File "C:\Users\12818\Desktop\software\QAT.Ultralytics\ultralytics\engine\trainer.py", line 452, in _do_train
preds = self.qat_model(batch['img'])
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\12818\AppData\Roaming\Python\Python312\site-packages\torch\fx\graph_module.py", line 822, in call_wrapped
return self._wrapped_call(self, *args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\12818\AppData\Roaming\Python\Python312\site-packages\torch\fx\graph_module.py", line 400, in call
raise e
File "C:\Users\12818\AppData\Roaming\Python\Python312\site-packages\torch\fx\graph_module.py", line 387, in call
return super(self.cls, obj).call(*args, **kwargs) # type: ignore[misc]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\12818\AppData\Roaming\Python\Python312\site-packages\torch\nn\modules\module.py", line 1739, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\12818\AppData\Roaming\Python\Python312\site-packages\torch\nn\modules\module.py", line 1750, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "<eval_with_key>.1167", line 2190, in forward
File "C:\Users\12818\AppData\Roaming\Python\Python312\site-packages\torch\nn\modules\module.py", line 1739, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\12818\AppData\Roaming\Python\Python312\site-packages\torch\nn\modules\module.py", line 1750, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\12818\AppData\Roaming\Python\Python312\site-packages\torch\ao\quantization\fake_quantize.py", line 408, in forward
return torch.fused_moving_avg_obs_fake_quant(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: zero_point must be between quant_min and quant_max.