mat1 and mat2 shapes cannot be multiplied for GRU

Question

I am creating a GRU to do some classification for a project, and I'm relatively new to Pytorch and implementing GRUs. I know similar questions like this one have been answered already but I can't seem to bring the same solution over to my own problem. I understand that there is an issue with the shape/order of my fc arrays but after trying to change things I can no longer see the trees for the wood. I would appreciate it if someone could point me in the right direction.

Below I have attached my code and the error. The datasets im using contain 24 features with a label in the 25th column.

# Imports
import pandas as pd
import numpy as np
import torch
import torchvision  # torch package for vision related things
import torch.nn.functional as F  # Parameterless functions, like (some) activation functions
import torchvision.datasets as datasets  # Standard datasets
import torchvision.transforms as transforms  # Transformations we can perform on our dataset for augmentation
from torch import optim  # For optimizers like SGD, Adam, etc.
from torch import nn  # All neural network modules
from torch.utils.data import Dataset, DataLoader  # Gives easier dataset managment by creating mini batches etc.
from tqdm import tqdm  # For a nice progress bar
from sklearn.preprocessing import StandardScaler

# Set device
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")

# Hyperparameters
input_size = 24
hidden_size = 128
num_layers = 1
num_classes = 2
sequence_length = 1
learning_rate = 0.005
batch_size = 8
num_epochs = 3

# Recurrent neural network with GRU (many-to-one)
class RNN_GRU(nn.Module):
    def __init__(self, input_size, hidden_size, num_layers, num_classes):
        super(RNN_GRU, self).__init__()
        self.hidden_size = hidden_size
        self.num_layers = num_layers
        self.gru = nn.GRU(input_size, hidden_size, num_layers, batch_first=True)
        self.fc = nn.Linear(hidden_size * sequence_length, num_classes)

    def forward(self, x):
        # Set initial hidden and cell states
        x = x.unsqueeze(0)
        h0 = torch.zeros(self.num_layers, x.size(0), self.hidden_size).to(device)

        # Forward propagate LSTM
        out, _ = self.gru(x, h0)
        out = out.reshape(out.shape[0], -1)

        # Decode the hidden state of the last time step
        out = self.fc(out)
        return out

class MyDataset(Dataset):
 
  def __init__(self,file_name):
    stats_df=pd.read_csv(file_name)
 
    x=stats_df.iloc[:,0:24].values
    y=stats_df.iloc[:,24].values
 
    self.x_train=torch.tensor(x,dtype=torch.float32)
    self.y_train=torch.tensor(y,dtype=torch.float32)
 
  def __len__(self):
    return len(self.y_train)
   
  def __getitem__(self,idx):
    return self.x_train[idx],self.y_train[idx]

nomDs=MyDataset("nomStats.csv")
atkDs=MyDataset("atkStats.csv")
train_loader=DataLoader(dataset=nomDs,batch_size=batch_size)
test_loader=DataLoader(dataset=atkDs,batch_size=batch_size)

# Initialize network (try out just using simple RNN, or GRU, and then compare with LSTM)
model = RNN_GRU(input_size, hidden_size, num_layers, num_classes).to(device)

# Loss and optimizer
criterion = nn.CrossEntropyLoss()
optimizer = optim.Adam(model.parameters(), lr=learning_rate)

# Train Network
for epoch in range(num_epochs):
    for batch_idx, (data, targets) in enumerate(tqdm(train_loader)):
        # Get data to cuda if possible
        data = data.to(device=device).squeeze(1)
        targets = targets.to(device=device)

        # forward
        scores = model(data)
        loss = criterion(scores, targets)

        # backward
        optimizer.zero_grad()
        loss.backward()

        # gradient descent update step/adam step
        optimizer.step()

# Check accuracy on training & test to see how good our model
def check_accuracy(loader, model):
    num_correct = 0
    num_samples = 0

    # Set model to eval
    model.eval()

    with torch.no_grad():
        for x, y in loader:
            x = x.to(device=device).squeeze(1)
            y = y.to(device=device)

            scores = model(x)
            _, predictions = scores.max(1)
            num_correct += (predictions == y).sum()
            num_samples += predictions.size(0)

    # Toggle model back to train
    model.train()
    return num_correct / num_samples


print(f"Accuracy on training set: {check_accuracy(train_loader, model)*100:2f}")
print(f"Accuracy on test set: {check_accuracy(test_loader, model)*100:.2f}")

Traceback (most recent call last):
  File "TESTGRU.py", line 87, in <module>
    scores = model(data)
  File "C:\Users\steph\anaconda3\envs\FYP\lib\site-packages\torch\nn\modules\module.py", line 889, in _call_impl
    result = self.forward(*input, **kwargs)
  File "TESTGRU.py", line 47, in forward
    out = self.fc(out)
  File "C:\Users\steph\anaconda3\envs\FYP\lib\site-packages\torch\nn\modules\module.py", line 889, in _call_impl
    result = self.forward(*input, **kwargs)
  File "C:\Users\steph\anaconda3\envs\FYP\lib\site-packages\torch\nn\modules\linear.py", line 94, in forward
    return F.linear(input, self.weight, self.bias)
  File "C:\Users\steph\anaconda3\envs\FYP\lib\site-packages\torch\nn\functional.py", line 1753, in linear
    return torch._C._nn.linear(input, weight, bias)
RuntimeError: mat1 and mat2 shapes cannot be multiplied (1x1024 and 128x2)

Answer 1

It seems like these lines

        # Forward propagate LSTM
        out, _ = self.gru(x, h0)
        out = out.reshape(out.shape[0], -1)

are the problem.

It appears that you only want to feed the hidden state of the last time step.

This could be read from the output in two ways:

1. If you want the output of all layers at the last time step, you should use the second return value of out, _ = self.gru(x, h0) not the first.

2. If you want to use just the last layer's output at the last time step (which seems to be the case), you should use out[:, -1, :]. With this change, you may not need the reshape operation.