Estou treinando uma auto-encoder
rede com Adam
otimizador (with amsgrad=True
) e MSE loss
para a tarefa de separação de fonte de áudio de canal único. Sempre que decai a taxa de aprendizado por um fator, a perda de rede aumenta bruscamente e depois diminui até a próxima queda na taxa de aprendizado.
Estou usando o Pytorch para implementação e treinamento de rede.
Following are my experimental setups:
Setup-1: NO learning rate decay, and
Using the same Adam optimizer for all epochs
Setup-2: NO learning rate decay, and
Creating a new Adam optimizer with same initial values every epoch
Setup-3: 0.25 decay in learning rate every 25 epochs, and
Creating a new Adam optimizer every epoch
Setup-4: 0.25 decay in learning rate every 25 epochs, and
NOT creating a new Adam optimizer every time rather
using PyTorch's "multiStepLR" and "ExponentialLR" decay scheduler
every 25 epochs
Estou obtendo resultados muito surpreendentes para as configurações 2, 3 e 4 e não consigo explicar nenhuma explicação. A seguir estão meus resultados:
Setup-1 Results:
Here I'm NOT decaying the learning rate and
I'm using the same Adam optimizer. So my results are as expected.
My loss decreases with more epochs.
Below is the loss plot this setup.
Trama-1:
optimizer = torch.optim.Adam(lr=m_lr,amsgrad=True, ...........)
for epoch in range(num_epochs):
running_loss = 0.0
for i in range(num_train):
train_input_tensor = ..........
train_label_tensor = ..........
optimizer.zero_grad()
pred_label_tensor = model(train_input_tensor)
loss = criterion(pred_label_tensor, train_label_tensor)
loss.backward()
optimizer.step()
running_loss += loss.item()
loss_history[m_lr].append(running_loss/num_train)
Setup-2 Results:
Here I'm NOT decaying the learning rate but every epoch I'm creating a new
Adam optimizer with the same initial parameters.
Here also results show similar behavior as Setup-1.
Because at every epoch a new Adam optimizer is created, so the calculated gradients
for each parameter should be lost, but it seems that this doesnot affect the
network learning. Can anyone please help on this?
Trama-2:
for epoch in range(num_epochs):
optimizer = torch.optim.Adam(lr=m_lr,amsgrad=True, ...........)
running_loss = 0.0
for i in range(num_train):
train_input_tensor = ..........
train_label_tensor = ..........
optimizer.zero_grad()
pred_label_tensor = model(train_input_tensor)
loss = criterion(pred_label_tensor, train_label_tensor)
loss.backward()
optimizer.step()
running_loss += loss.item()
loss_history[m_lr].append(running_loss/num_train)
Setup-3 Results:
As can be seen from the results in below plot,
my loss jumps every time I decay the learning rate. This is a weird behavior.
If it was happening due to the fact that I'm creating a new Adam
optimizer every epoch then, it should have happened in Setup #1, #2 as well.
And if it is happening due to the creation of a new Adam optimizer with a new
learning rate (alpha) every 25 epochs, then the results of Setup #4 below also
denies such correlation.
Trama-3:
decay_rate = 0.25
for epoch in range(num_epochs):
optimizer = torch.optim.Adam(lr=m_lr,amsgrad=True, ...........)
if epoch % 25 == 0 and epoch != 0:
lr *= decay_rate # decay the learning rate
running_loss = 0.0
for i in range(num_train):
train_input_tensor = ..........
train_label_tensor = ..........
optimizer.zero_grad()
pred_label_tensor = model(train_input_tensor)
loss = criterion(pred_label_tensor, train_label_tensor)
loss.backward()
optimizer.step()
running_loss += loss.item()
loss_history[m_lr].append(running_loss/num_train)
Setup-4 Results:
In this setup, I'm using Pytorch's learning-rate-decay scheduler (multiStepLR)
which decays the learning rate every 25 epochs by 0.25.
Here also, the loss jumps everytime the learning rate is decayed.
Como sugerido por @Dennis nos comentários abaixo, tentei com ambas ReLU
e 1e-02 leakyReLU
não - linearidades. Mas, os resultados parecem se comportar de maneira semelhante e a perda primeiro diminui, depois aumenta e depois satura em um valor mais alto do que o que eu obteria sem diminuir a taxa de aprendizado.
O gráfico 4 mostra os resultados.
Trama-4:
scheduler = torch.optim.lr_scheduler.MultiStepLR(optimizer=optimizer, milestones=[25,50,75], gamma=0.25)
scheduler = torch.optim.lr_scheduler.ExponentialLR(optimizer=optimizer, gamma=0.95)
scheduler = ......... # defined above
optimizer = torch.optim.Adam(lr=m_lr,amsgrad=True, ...........)
for epoch in range(num_epochs):
scheduler.step()
running_loss = 0.0
for i in range(num_train):
train_input_tensor = ..........
train_label_tensor = ..........
optimizer.zero_grad()
pred_label_tensor = model(train_input_tensor)
loss = criterion(pred_label_tensor, train_label_tensor)
loss.backward()
optimizer.step()
running_loss += loss.item()
loss_history[m_lr].append(running_loss/num_train)
EDITAR% S:
- Conforme sugerido nos comentários e respostas abaixo, fiz alterações no meu código e treinei o modelo. Eu adicionei o código e os gráficos para o mesmo.
- Tentei com vários
lr_scheduler
inPyTorch (multiStepLR, ExponentialLR)
e plotagens para o mesmo estão listadosSetup-4
como sugerido por @Dennis nos comentários abaixo. - Tentando com leakyReLU como sugerido por @Dennis nos comentários.
Qualquer ajuda. obrigado