RuntimeError: expected scalar type Long but found Float

[garvpatidar04]

epochs = 200

for epoch in range(epochs):
model_3.train()

y_logit = model_3(X_train1)

loss = loss_fn1(y_logit, y_train1)

optimizer.zero_grad()

loss.backward()

optimizer.step()

model_3.eval()

with torch.inference_mode():
y_logit_t = model_3(X_test1)

loss_test = loss_fn1(y_logit_t, y_test1)

if epoch%40==0:
  print(f'Epoch: {epoch} | training loss: {loss} | test loss: {loss_test}') 
  
above is the code for the training loop and I am getting a type error

and y_train1 is like this

and y_logits is like this

These are the optimizer and loss function

I have tried to change the dtype of y_train1 and y_logits, but it didn't work

Can anyone help me to resolve the issue here, It would be a great help to me

[LuluW8071]

Typical datatype error
Just change ytrain1 to long dtype. This is done for labels since labels should be fixed value not floating points.

Paste below snippet above your training loop code

y_train1 = y_train1.type(torch.LongTensor)

[garvpatidar04]

Thank you, it did solve the problem, earlier I tried to change the type to torch.long and that didn't work because I changed it to inside the loop 😅.