When trying to run my classification model with pytorch I get the error:
RuntimeError: expected scalar type Long but found Float
Which is weird since I looked at line where the error is supposed to be x = self.linear1(x) and checked with the debugger the type of x and its of type LongTensor.
I also tried adding .long() to x which doesnt seems to help either.
Here`s the full error message:
Validation sanity check: 0%| | 0/2 [00:00<?, ?it/s]Traceback (most recent call last):
File "E:/Programming/workspace_python/LOL-Match-Outcome-Predictor/code/leaguePreditctSimple_train.py", line 287, in <module>
trainer.fit(net)
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\pytorch_lightning\trainer\trainer.py", line 552, in fit
self._run(model)
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\pytorch_lightning\trainer\trainer.py", line 917, in _run
self._dispatch()
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\pytorch_lightning\trainer\trainer.py", line 985, in _dispatch
self.accelerator.start_training(self)
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\pytorch_lightning\accelerators\accelerator.py", line 92, in start_training
self.training_type_plugin.start_training(trainer)
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\pytorch_lightning\plugins\training_type\training_type_plugin.py", line 161, in start_training
self._results = trainer.run_stage()
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\pytorch_lightning\trainer\trainer.py", line 995, in run_stage
return self._run_train()
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\pytorch_lightning\trainer\trainer.py", line 1030, in _run_train
self._run_sanity_check(self.lightning_module)
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\pytorch_lightning\trainer\trainer.py", line 1114, in _run_sanity_check
self._evaluation_loop.run()
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\pytorch_lightning\loops\base.py", line 111, in run
self.advance(*args, **kwargs)
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\pytorch_lightning\loops\dataloader\evaluation_loop.py", line 111, in advance
dataloader_iter, self.current_dataloader_idx, dl_max_batches, self.num_dataloaders
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\pytorch_lightning\loops\base.py", line 111, in run
self.advance(*args, **kwargs)
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\pytorch_lightning\loops\epoch\evaluation_epoch_loop.py", line 110, in advance
output = self.evaluation_step(batch, batch_idx, dataloader_idx)
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\pytorch_lightning\loops\epoch\evaluation_epoch_loop.py", line 154, in evaluation_step
output = self.trainer.accelerator.validation_step(step_kwargs)
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\pytorch_lightning\accelerators\accelerator.py", line 211, in validation_step
return self.training_type_plugin.validation_step(*step_kwargs.values())
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\pytorch_lightning\plugins\training_type\training_type_plugin.py", line 178, in validation_step
return self.model.validation_step(*args, **kwargs)
File "E:/Programming/workspace_python/LOL-Match-Outcome-Predictor/code/leaguePreditctSimple_train.py", line 170, in validation_step
y_hat = self(x)
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl
return forward_call(*input, **kwargs)
File "E:/Programming/workspace_python/LOL-Match-Outcome-Predictor/code/leaguePreditctSimple_train.py", line 151, in forward
x = self.linear1(x)
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl
return forward_call(*input, **kwargs)
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\torch\nn\modules\container.py", line 139, in forward
input = module(input)
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl
return forward_call(*input, **kwargs)
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\torch\nn\modules\linear.py", line 96, in forward
return F.linear(input, self.weight, self.bias)
File "E:\Programming\workspace_python\LOL-Match-Outcome-Predictor\venv\lib\site-packages\torch\nn\functional.py", line 1847, in linear
return torch._C._nn.linear(input, weight, bias)
RuntimeError: expected scalar type Long but found Float
Here`s the network code:
class NN(pl.LightningModule):
def __init__(self, hparams, experiment_name):
super().__init__()
# self.hparams = hparams # Depreceated in newer versions
self.save_hyperparameters(hparams)
self.checkpoint_folder = "LeaguePredictCheckpoints/"
self.experiment_name = experiment_name
# creating checkpoint folder
dirpath = Path(self.checkpoint_folder)
if not dirpath.exists():
os.makedirs(dirpath, 0o755)
self.linear1 = nn.Sequential(
nn.Linear(in_features=1480, out_features=50), # "*10" because we have 10 players
nn.PReLU(),
)
self.linear2 = nn.Sequential(
nn.Linear(in_features=50, out_features=20),
# nn.BatchNorm1d(num_features=20), # Batchnorm only in hidden layers?
nn.PReLU()
)
self.linear3 = nn.Sequential(
nn.Linear(in_features=20, out_features=2),
nn.Sigmoid()
)
def forward(self, x):
x = self.linear1(x)
x = self.linear2(x)
x = self.linear3(x)
return x
def training_step(self, batch, batch_idx):
x, y = batch
y_hat = self(x)
loss = self.loss(y_hat, y)
# calculate accuracy
y_hat = torch.argmax(y_hat, dim=1)
y = torch.argmax(y, dim=1)
acc_train = torch.tensor(torch.sum(y == y_hat).item() / (len(y) * 1.0))
return {"loss": loss, "acc_train": acc_train}
def validation_step(self, batch, batch_idx):
x, y = batch
y_hat = self(x)
loss = self.loss(y_hat, y)
# calculate accuracy
y_hat = torch.argmax(y_hat, dim=1)
y = torch.argmax(y, dim=1)
acc_val = torch.tensor(torch.sum(y == y_hat).item() / (len(y) * 1.0))
return {"val_loss": loss, "acc_val": acc_val}
def validation_epoch_end(self, outputs):
# average loss per epoch
avg_loss = torch.stack([x['val_loss'] for x in outputs]).mean()
self.log("avg_epoch_val_loss", avg_loss)
# average accuracy per epoch
avg_acc = torch.stack([x['acc_val'] for x in outputs]).mean()
self.log("avg_epoch_val_acc", avg_acc)
return {'val_loss': avg_loss}
def training_epoch_end(self, outputs):
# average loss per epoch
avg_loss = torch.stack([x['loss'] for x in outputs]).mean()
self.log("avg_epoch_train_loss", avg_loss)
# average accuracy per epoch
avg_acc = torch.stack([x['acc_train'] for x in outputs]).mean()
self.log("avg_epoch_train_acc", avg_acc)
def loss(self, input, target):
return F.binary_cross_entropy(input.float(), target.float())
def configure_optimizers(self):
return torch.optim.Adam(self.parameters(), lr=self.hparams.LR)
def train_dataloader(self):
train_dataset = LaegueDataset_train("../train_dataset_location.txt", "../championidtable_location.txt", self.hparams.CHUNKSIZE)
return DataLoader(train_dataset, num_workers=self.hparams.NUMWORK, batch_size=self.hparams.BATCHSIZE)
def val_dataloader(self):
val_dataset = LaegueDataset_train("../test_dataset_location.txt", "../championidtable_location.txt", self.hparams.CHUNKSIZE)
return DataLoader(val_dataset, num_workers=self.hparams.NUMWORK, batch_size=self.hparams.BATCHSIZE)
def on_epoch_end(self) -> None:
if self.current_epoch % self.hparams.SAVE_MODEL_EVERY_EPOCH == 0:
trainer.save_checkpoint(
self.checkpoint_folder + "/" + self.experiment_name + "_epoch_" + str(self.current_epoch) + ".ckpt")
comet_logger.experiment.log_asset_folder(self.checkpoint_folder, step=self.current_epoch)
# Deleting the folder where we saved the model so that we dont upload a thing twice
dirpath = Path(self.checkpoint_folder)
if dirpath.exists() and dirpath.is_dir():
shutil.rmtree(dirpath)
# creating checkpoint folder
access_rights = 0o755
os.makedirs(dirpath, access_rights)
def on_train_end(self):
trainer.save_checkpoint(
self.checkpoint_folder + "/" + self.experiment_name + "_epoch_" + str(self.current_epoch) + ".ckpt")
comet_logger.experiment.log_asset_folder(self.checkpoint_folder, step=self.current_epoch)