in vscode ipynb
Using val_dataloaders will print the </> symbol for the number of epochs.
Please let me know how to resolve this.
Hey @kyounoii
This is not a lot of information. I’m not sure how to help here. Are you referring to a behavior in the progress bar?
I’m not talking about the progress bar. It is a number of symbols that are output under it.
I am attaching a screenshot and the full program.
I am using the latest version of lightning.
The symbol in question is not printed in *.py. It is output in *.ipynb.
import numpy as np
import pandas as pd
df=pd.read_csv(‘housing.csv’)
df.head()
x=df.drop(columns=‘y’).values
t=df[‘y’].values
import lightning
import torch
x=torch.tensor(x,dtype=torch.float32)
t=torch.tensor(t,dtype=torch.float32)
import torch.utils.data
dataset=torch.utils.data.TensorDataset(x,t)
torch.manual_seed(0)
n_train=int(len(dataset)*0.6)
n_val=int(len(dataset)*0.2)
n_test=len(dataset)-n_train-n_val
train,val,test =torch.utils.data.random_split(dataset,[n_train,n_val,n_test])
import lightning as pl
from typing import Any
from pytorch_lightning.utilities.types import STEP_OUTPUT, TRAIN_DATALOADERS
class TrainNet(pl.LightningModule):
def train_dataloader(self):
return torch.utils.data.DataLoader(train, batch_size=self.batch_size, shuffle=True)
def training_step(self, batch, batch_nb):
x,t=batch
y=self.forward(x)
loss=self.lossfun(y,t)
self.log('kaiki_loss',loss)
return loss
class ValidationNet(pl.LightningModule):
def val_dataloader(self):
return torch.utils.data.DataLoader(val, batch_size=self.batch_size)
def validation_step(self, batch, batch_nb):
x,t=batch
y=self.forward(x)
loss=self.lossfun(y,t)
self.log('val_kaiki_loss',loss)
# return loss
class TestNet(pl.LightningModule):
def test_dataloader(self):
return torch.utils.data.DataLoader(test, batch_size=self.batch_size)
def test_step(self, batch, batch_nb):
x,t=batch
y=self.forward(x)
loss=self.lossfun(y,t)
self.log(‘test_kaiki_loss’,loss)
return loss
class Net(TrainNet,ValidationNet,TestNet):
def init(self,input_size=13,hidden_size=15,output_size=1,batch_size=10):
super().init()
self.fc1=torch.nn.Linear(input_size,hidden_size)
self.fc2=torch.nn.Linear(hidden_size,output_size)
self.batch_size=batch_size
def forward(self, x):
x=self.fc1(x)
x=F.relu(x)
x=self.fc2(x)
return x
def lossfun(self,y,t):
return F.mse_loss(y,t)
def configure_optimizers(self):
return torch.optim.SGD(self.parameters(),lr=1e-1)
torch.manual_seed(0)
net=Net()
trainer=pl.Trainer(max_epochs=100)
trainer.fit(net)