下面的代码在单个 GPU 上工作，但在使用多个 GPU 时抛出错误 RuntimeError: grad can be implicitly created only for scalar outputs
代码
def forward(
self,
input_ids,
attention_mask=None,
decoder_input_ids=None,
decoder_attention_mask=None,
lm_labels=None
):
return self.model(
input_ids,
attention_mask=attention_mask,
decoder_input_ids=decoder_input_ids,
decoder_attention_mask=decoder_attention_mask,
labels=lm_labels,
)
def _step(self, batch):
lm_labels = batch["target_ids"]
# lm_labels[lm_labels[:, :] == self.tokenizer.pad_token_id] = -100
outputs = self(
input_ids=batch["source_ids"],
attention_mask=batch["source_mask"],
lm_labels=lm_labels,
decoder_attention_mask=batch['target_mask']
)
loss = outputs[0]
return loss
def training_step(self, batch, batch_idx):
loss = self._step(batch)
return {"loss": loss}
损失值是一个缩放器： tensor(12.8875, device='cuda:1', grad_fn=NllLossBackward) 这个错误背后的原因是什么？
回溯（最后一次调用）：文件“training_trial.py”，第 390 行，在 trainer.fit(model) 文件“/home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/pytorch_lightning /trainer/trainer.py”，第 510 行，适合结果 = self.accelerator_backend.train() 文件“/home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/pytorch_lightning/accelerators/加速器.py”，第 57 行，在火车中返回 self.train_or_test() 文件“/home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/pytorch_lightning/accelerators/accelerator.py”，行74，在train_or_test results = self.trainer.train() 文件“/home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/pytorch_lightning/trainer/trainer.py”，第561行，在训练自己。train_loop.run_training_epoch() 文件“/home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/pytorch_lightning/trainer/training_loop.py”，第 549 行，在 run_training_epoch batch_output = self.run_training_batch(batch ，batch_idx，dataloader_idx）文件“/home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/pytorch_lightning/trainer/training_loop.py”，第704行，在run_training_batch self.optimizer_step（优化器，opt_idx ，batch_idx，train_step_and_backward_closure）文件“/home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/pytorch_lightning/trainer/training_loop.py”，第490行，在optimizer_step using_lbfgs=is_lbfgs，文件“/主页/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/pytorch_lightning/core/lightning。py”，第 1296 行，在 optimizer_step optimizer.step(closure=optimizer_closure) 文件“/home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/pytorch_lightning/core/optimizer.py”，行286、在步骤self.__optimizer_step(*args,closure=closure,profiler_name=profiler_name,**kwargs)文件“/home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/pytorch_lightning/core /optimizer.py"，第 144 行，在 _7/site-packages/pytorch_lightning/core/optimizer.py”，第 144 行，在 _7/site-packages/pytorch_lightning/core/optimizer.py”，第 144 行，在 _optimizer_step optimizer.step(closure=closure, *args, **kwargs) 文件“/home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/torch/optim/lr_scheduler.py”，行67、在 wrapper 中返回 Wrapped(*args, **kwargs) 文件“/home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/transformers/optimization.py”，第 318 行，步骤loss =closure() 文件“/home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/pytorch_lightning/trainer/training_loop.py”，第 699 行，在 train_step_and_backward_closure self.trainer.hiddens 文件中“ /home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/pytorch_lightning/trainer/training_loop.py”，第 802 行，training_step_and_backward self.backward(result, optimizer, opt_idx) 文件”/home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/pytorch_lightning/trainer/training_loop.py”，第 829 行，后向 result.closure_loss，优化器，opt_idx，*args，**kwargs文件“/home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/pytorch_lightning/accelerators/accelerator.py”，第 109 行，后向 model.backward（closure_loss，优化器，opt_idx，*args , **kwargs) 文件“/home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/pytorch_lightning/core/lightning.py”，第 1162 行，后向 loss.backward(*args, **kwargs）文件“/home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/torch/tensor.py”，第 221 行，后向 torch.autograd.backward（self，gradient，保留图，创建图）文件“/home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/torch/autograd/init .py”，第 126 行，向后 grad_tensors = make_grads(tensors, grad_tensors ) 文件“/home/nvarshn2/.conda/envs/pytorch_lightning_new_env/lib/python3.7/site-packages/torch/autograd/init .py ” ，第 50 行，在 _make_grads raise RuntimeError("grad can be implicitly created only for scalar outputs") RuntimeError: grad can be implicitly created only for scalar outputs