Added working model-parallel tutorial

accre · Dec 23, 2023 · 7a15c87 · 7a15c87
1 parent dfb208d
commit 7a15c87
Show file tree

Hide file tree

Showing 11 changed files with 5 additions and 6 deletions.
diff --git a/.train_pytorch_modelParallel.py.swp b/.train_pytorch_modelParallel.py.swp
diff --git a/data/FashionMNIST/raw/t10k-images-idx3-ubyte b/data/FashionMNIST/raw/t10k-images-idx3-ubyte
diff --git a/data/FashionMNIST/raw/t10k-images-idx3-ubyte.gz b/data/FashionMNIST/raw/t10k-images-idx3-ubyte.gz
diff --git a/data/FashionMNIST/raw/t10k-labels-idx1-ubyte b/data/FashionMNIST/raw/t10k-labels-idx1-ubyte
diff --git a/data/FashionMNIST/raw/t10k-labels-idx1-ubyte.gz b/data/FashionMNIST/raw/t10k-labels-idx1-ubyte.gz
diff --git a/data/FashionMNIST/raw/train-images-idx3-ubyte b/data/FashionMNIST/raw/train-images-idx3-ubyte
diff --git a/data/FashionMNIST/raw/train-images-idx3-ubyte.gz b/data/FashionMNIST/raw/train-images-idx3-ubyte.gz
diff --git a/data/FashionMNIST/raw/train-labels-idx1-ubyte b/data/FashionMNIST/raw/train-labels-idx1-ubyte
diff --git a/data/FashionMNIST/raw/train-labels-idx1-ubyte.gz b/data/FashionMNIST/raw/train-labels-idx1-ubyte.gz
diff --git a/train_pytorch_modelParallel.py b/train_pytorch_modelParallel.py
@@ -42,7 +42,7 @@ def train(model, train_loader, loss_function, optimizer, num_epochs):
 
         for i ,(images,labels) in enumerate(train_loader):
             images = torch.div(images, 255.)
-            images, labels = images.to(device), labels.to(device)
+#            images, labels = images.to(device), labels.to(device)
 
             optimizer.zero_grad()
             outputs = model(images)
@@ -70,9 +70,9 @@ def train(model, train_loader, loss_function, optimizer, num_epochs):
 if not torch.cuda.is_available():
   sys.exit("A minimum of 2 GPUs must be available to train this model.")
 
-print("Training on device: ", device)
+#print("Training on device: ", device)
 my_net = SeqNet(input_size, hidden_size1, output_size)
-my_net = my_net.to(device)
+#my_net = my_net.to(device)
 
 
 optimizer = torch.optim.Adam( my_net.parameters(), lr=lr) 
@@ -86,6 +86,7 @@ def train(model, train_loader, loss_function, optimizer, num_epochs):
 
 train(my_net, fmnist_train_loader, loss_function, optimizer, num_epochs) 
 
+"""
 correct = 0
 total = 0
 for images,labels in fmnist_test_loader:
@@ -96,5 +97,5 @@ def train(model, train_loader, loss_function, optimizer, num_epochs):
   _, predicted = torch.max(output,1)
   correct += (predicted == labels).sum()
   total += labels.size(0)
-
 print('Accuracy of the model: %.3f %%' %((100*correct)/(total+1)))
+"""
diff --git a/tutorials/model-parallel.md b/tutorials/model-parallel.md
@@ -50,7 +50,6 @@ def train(model, train_loader, loss_function, optimizer, num_epochs):
 
         for i ,(images,labels) in enumerate(train_loader):
             images = torch.div(images, 255.)
-            images, labels = images.to(device), labels.to(device)
 
             optimizer.zero_grad()
             outputs = model(images)
@@ -81,7 +80,6 @@ lr = 0.01
 if not torch.cuda.is_available():
   sys.exit("A minimum of 2 GPUs must be available to train this model.")
 
-print("Training on device: ", device)
 my_net = SeqNet(input_size, hidden_size1, output_size)
 
 optimizer = torch.optim.Adam( my_net.parameters(), lr=lr)