why i can't run the code of 01. PyTorch Workflow Fundamentals

[axuanyaa]

My error

RuntimeError: Trying to backward through the graph a second time (or directly access saved tensors after they have already been freed). Saved intermediate values of the graph are freed when you call .backward() or autograd.grad(). Specify retain_graph=True if you need to backward through the graph a second time or if you need to access saved tensors after calling backward.

My code

training loop and testing loop

RANDOM_SEED = 42
torch.manual_seed(RANDOM_SEED)

epochs = 100

put the data to printed device

X_train_1 = X_train_1.to(device)
y_train_1 = y_train_1.to(device)
X_test_1 = X_test_1.to(device)
y_test_1 = y_test_1.to(device)

loss_train_1 = []
loss_test_1 = []

training loop

for epoch in range(epochs):

change to train mode

model_1.train()

forward pass

y_train_1_pred = model_1(X_train_1)

calculate loss function

loss_1 = loss_fn_1(y_train_1_pred, y_train_1)

set parameters to zero_grad

optimizer_1.zero_grad()

backward

loss_1.backward()

update

optimizer_1.step()

change to test mode

model_1.eval()

with no_grad

with torch.inference_mode():
# forward pass
y_test_1_pred = model_1(X_test_1)

# calculate the loss
loss_1_test = loss_fn_1(y_test_1_pred, y_test_1)

store the loss

if epoch % 10 == 0:
loss_train_1.append(loss_1)
loss_test_1.append(loss_1_test)
print(f"epoch:{epoch}|train loss:{loss_1}|test loss:{loss_1_test}")

What I've tried so far

i dont konw why ,and after searching for google and gpt ,i didnt get the answer, plsease help me