RuntimeError: CUDA error: device-side assert triggered

当我运行使用集合预测网络在GPU服务器上联合提取实体关系的代码时，我遇到了这个错误:
/pytorch/aten/src/ATen/native/cuda/Loss.cu:247: nll_loss_forward_reduce_cuda_kernel_2d: block: [0,0,0], thread: [0,0,0] Assertion t >= 0 && t < n_classes failed.
Traceback (most recent call last):
File "/root/miniconda3/lib/python3.8/runpy.py", line 194, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/root/miniconda3/lib/python3.8/runpy.py", line 87, in _run_code
exec(code, run_globals)
File "/root/autodl-tmp/SPN4RE/Nr_Partial_ch_SPN4RE-main/main.py", line 102, in
trainer.train_model()
File "/root/autodl-tmp/SPN4RE/Nr_Partial_ch_SPN4RE-main/trainer/trainer.py", line 77, in train_model
loss, _ = self.model(input_ids, attention_mask, targets)
File "/root/miniconda3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "/root/autodl-tmp/SPN4RE/Nr_Partial_ch_SPN4RE-main/models/setpred4RE.py", line 31, in forward
loss = self.criterion(outputs, targets)
File "/root/miniconda3/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "/root/autodl-tmp/SPN4RE/Nr_Partial_ch_SPN4RE-main/models/set_criterion.py", line 46, in forward
losses.update(self.get_loss(loss, outputs, targets, indices))
File "/root/autodl-tmp/SPN4RE/Nr_Partial_ch_SPN4RE-main/models/set_criterion.py", line 96, in get_loss
return loss_map[loss](outputs, targets, indices, **kwargs)
File "/root/autodl-tmp/SPN4RE/Nr_Partial_ch_SPN4RE-main/models/set_criterion.py", line 113, in entity_loss
head_start_loss = F.cross_entropy(selected_pred_head_start, target_head_start)
File "/root/miniconda3/lib/python3.8/site-packages/torch/nn/functional.py", line 2846, in cross_entropy
return torch._C._nn.cross_entropy_loss(input, target, weight, _Reduction.get_enum(reduction), ignore_index, label_smoothing)
RuntimeError: CUDA error: device-side assert triggered
请问应该怎么解决？