Skip to content

Fix torch.load device handling and remove deprecated scheduler params #134

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
Aryanjstar wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Fix torch.load device handling and remove deprecated scheduler params #134

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 3 additions & 8 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Anomaly_Detection/Model_II/AAE/test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -47,14 +47,9 @@ def final_test_epoch(model, dataloader,criterion):
decoder = Decoder().to(device)
disc = Discriminator().to(device)

if device != 'cpu':
encoder.load_state_dict(torch.load(MODEL_PATH + 'encoder.pth'))
decoder.load_state_dict(torch.load(MODEL_PATH + 'decoder.pth'))
disc.load_state_dict(torch.load(MODEL_PATH + 'discriminator.pth'))
else:
encoder.load_state_dict(torch.load(MODEL_PATH + 'encoder.pth', map_location = torch.device('cpu')))
decoder.load_state_dict(torch.load(MODEL_PATH + 'decoder.pth', map_location = torch.device('cpu')))
disc.load_state_dict(torch.load(MODEL_PATH + 'discriminator.pth', map_location = torch.device('cpu')))
encoder.load_state_dict(torch.load(MODEL_PATH + 'encoder.pth', map_location=device, weights_only=True))
decoder.load_state_dict(torch.load(MODEL_PATH + 'decoder.pth', map_location=device, weights_only=True))
disc.load_state_dict(torch.load(MODEL_PATH + 'discriminator.pth', map_location=device, weights_only=True))

model = {'enc':encoder,
'dec': decoder,
Expand Down
12 changes: 3 additions & 9 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Anomaly_Detection/Model_II/AAE/train.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -201,15 +201,9 @@ def fit_model(model, checkpoint_path):
disc = Discriminator().to(device)

if LOAD_PRETRAINED_MODEL:
if device != 'cpu':
encoder.load_state_dict(torch.load(MODEL_PATH + 'encoder.pth'))
decoder.load_state_dict(torch.load(MODEL_PATH + 'decoder.pth'))
disc.load_state_dict(torch.load(MODEL_PATH + 'discriminator.pth'))

else:
encoder.load_state_dict(torch.load(MODEL_PATH + 'encoder.pth', map_location = torch.device('cpu')))
decoder.load_state_dict(torch.load(MODEL_PATH + 'decoder.pth', map_location = torch.device('cpu')))
disc.load_state_dict(torch.load(MODEL_PATH + 'discriminator.pth', map_location = torch.device('cpu')))
encoder.load_state_dict(torch.load(MODEL_PATH + 'encoder.pth', map_location=device, weights_only=True))
decoder.load_state_dict(torch.load(MODEL_PATH + 'decoder.pth', map_location=device, weights_only=True))
disc.load_state_dict(torch.load(MODEL_PATH + 'discriminator.pth', map_location=device, weights_only=True))

model = {'enc':encoder,
'dec': decoder,
Expand Down
5 changes: 1 addition & 4 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Anomaly_Detection/Model_II/AE/test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -39,10 +39,7 @@ def final_test_epoch(model, dataloader,criterion):
set_seed(7)

model = Autoencoder().to(device)
if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

full_test_loader, class_map = create_full_test_dataloader(FULL_TEST_DATA_PATH, test_transforms, BATCH_SIZE)
criterion = nn.MSELoss()
Expand Down
5 changes: 1 addition & 4 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Anomaly_Detection/Model_II/AE/train.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -148,10 +148,7 @@ def fit_model(model, checkpoint_path):
model = Autoencoder().to(device)

if LOAD_PRETRAINED_MODEL:
if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

model, loss_dict = fit_model(model, checkpoint_path = MODEL_PATH)

Expand Down
5 changes: 1 addition & 4 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Anomaly_Detection/Model_II/VAE/model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -93,8 +93,5 @@ def forward(self, x):
encoder = Encoder()
decoder = Decoder()
model = VAE(encoder, decoder)
if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))
print(model)
5 changes: 1 addition & 4 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Anomaly_Detection/Model_II/VAE/test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -40,10 +40,7 @@ def final_test_epoch(model, dataloader,criterion):
encoder = Encoder().to(device)
decoder = Decoder().to(device)
model = VAE(encoder, decoder).to(device)
if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

full_test_loader, class_map = create_full_test_dataloader(FULL_TEST_DATA_PATH, test_transforms, BATCH_SIZE)
criterion = nn.MSELoss()
Expand Down
7 changes: 2 additions & 5 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Anomaly_Detection/Model_II/VAE/train.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -121,7 +121,7 @@ def plot_ae_outputs(model, dataloader, n = 10):
def fit_model(model, checkpoint_path):

optimizer = optim.Adam(model.parameters(), lr = LEARNING_RATE)
scheduler = optim.lr_scheduler.OneCycleLR(optimizer, LEARNING_RATE, epochs = EPOCHS, steps_per_epoch = len(train_loader), verbose = False)
scheduler = optim.lr_scheduler.OneCycleLR(optimizer, LEARNING_RATE, epochs=EPOCHS, steps_per_epoch=len(train_loader))
criterion = nn.MSELoss()

loss_dict = {'train_loss':[],'val_loss':[], 'train_kld_loss':[], 'val_kld_loss':[]}
Expand Down Expand Up @@ -170,10 +170,7 @@ def fit_model(model, checkpoint_path):
model = VAE(encoder, decoder).to(device)

if LOAD_PRETRAINED_MODEL:
if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

model, loss_dict = fit_model(model, checkpoint_path = MODEL_PATH)

Expand Down
11 changes: 3 additions & 8 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Anomaly_Detection/Model_III/AAE/test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -47,14 +47,9 @@ def final_test_epoch(model, dataloader,criterion):
decoder = Decoder().to(device)
disc = Discriminator().to(device)

if device != 'cpu':
encoder.load_state_dict(torch.load(MODEL_PATH + 'encoder.pth'))
decoder.load_state_dict(torch.load(MODEL_PATH + 'decoder.pth'))
disc.load_state_dict(torch.load(MODEL_PATH + 'discriminator.pth'))
else:
encoder.load_state_dict(torch.load(MODEL_PATH + 'encoder.pth', map_location = torch.device('cpu')))
decoder.load_state_dict(torch.load(MODEL_PATH + 'decoder.pth', map_location = torch.device('cpu')))
disc.load_state_dict(torch.load(MODEL_PATH + 'discriminator.pth', map_location = torch.device('cpu')))
encoder.load_state_dict(torch.load(MODEL_PATH + 'encoder.pth', map_location=device, weights_only=True))
decoder.load_state_dict(torch.load(MODEL_PATH + 'decoder.pth', map_location=device, weights_only=True))
disc.load_state_dict(torch.load(MODEL_PATH + 'discriminator.pth', map_location=device, weights_only=True))

model = {'enc':encoder,
'dec': decoder,
Expand Down
11 changes: 3 additions & 8 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Anomaly_Detection/Model_III/AAE/train.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -203,14 +203,9 @@ def fit_model(model, checkpoint_path):
disc = Discriminator().to(device)

if LOAD_PRETRAINED_MODEL:
if device != 'cpu':
encoder.load_state_dict(torch.load(MODEL_PATH + 'encoder.pth'))
decoder.load_state_dict(torch.load(MODEL_PATH + 'decoder.pth'))
disc.load_state_dict(torch.load(MODEL_PATH + 'discriminator.pth'))
else:
encoder.load_state_dict(torch.load(MODEL_PATH + 'encoder.pth', map_location = torch.device('cpu')))
decoder.load_state_dict(torch.load(MODEL_PATH + 'decoder.pth', map_location = torch.device('cpu')))
disc.load_state_dict(torch.load(MODEL_PATH + 'discriminator.pth', map_location = torch.device('cpu')))
encoder.load_state_dict(torch.load(MODEL_PATH + 'encoder.pth', map_location=device, weights_only=True))
decoder.load_state_dict(torch.load(MODEL_PATH + 'decoder.pth', map_location=device, weights_only=True))
disc.load_state_dict(torch.load(MODEL_PATH + 'discriminator.pth', map_location=device, weights_only=True))

model = {'enc':encoder,
'dec': decoder,
Expand Down
5 changes: 1 addition & 4 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Anomaly_Detection/Model_III/AE/test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -39,10 +39,7 @@ def final_test_epoch(model, dataloader,criterion):
set_seed(7)

model = Autoencoder().to(device)
if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

full_test_loader, class_map = create_full_test_dataloader(FULL_TEST_DATA_PATH, test_transforms, BATCH_SIZE)
criterion = nn.MSELoss()
Expand Down
5 changes: 1 addition & 4 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Anomaly_Detection/Model_III/AE/train.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -147,10 +147,7 @@ def fit_model(model, checkpoint_path):
model = Autoencoder().to(device)

if LOAD_PRETRAINED_MODEL:
if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

model, loss_dict = fit_model(model, checkpoint_path = MODEL_PATH)

Expand Down
5 changes: 1 addition & 4 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Anomaly_Detection/Model_III/VAE/model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -93,8 +93,5 @@ def forward(self, x):
encoder = Encoder()
decoder = Decoder()
model = VAE(encoder, decoder)
if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))
print(model)
5 changes: 1 addition & 4 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Anomaly_Detection/Model_III/VAE/test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -40,10 +40,7 @@ def final_test_epoch(model, dataloader,criterion):
encoder = Encoder().to(device)
decoder = Decoder().to(device)
model = VAE(encoder, decoder).to(device)
if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

full_test_loader, class_map = create_full_test_dataloader(FULL_TEST_DATA_PATH, test_transforms, BATCH_SIZE)
criterion = nn.MSELoss()
Expand Down
7 changes: 2 additions & 5 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Anomaly_Detection/Model_III/VAE/train.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -122,7 +122,7 @@ def fit_model(model,checkpoint_path):


optimizer = optim.Adam(model.parameters(), lr = LEARNING_RATE)
scheduler = optim.lr_scheduler.OneCycleLR(optimizer, LEARNING_RATE, epochs = EPOCHS, steps_per_epoch = len(train_loader), verbose = False)
scheduler = optim.lr_scheduler.OneCycleLR(optimizer, LEARNING_RATE, epochs=EPOCHS, steps_per_epoch=len(train_loader))
criterion = nn.MSELoss()

loss_dict = {'train_loss':[],'val_loss':[], 'train_kld_loss':[], 'val_kld_loss':[]}
Expand Down Expand Up @@ -171,10 +171,7 @@ def fit_model(model,checkpoint_path):
model = VAE(encoder, decoder).to(device)

if LOAD_PRETRAINED_MODEL:
if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

model, loss_dict = fit_model(model, checkpoint_path = MODEL_PATH)

Expand Down
5 changes: 1 addition & 4 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Classification/Model_I/test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -63,10 +63,7 @@ def flatten_list(x):
criterion = nn.CrossEntropyLoss()
model = EffNetB2_backbone_model().to(device)

if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

_, _, test_loader, class_map = create_data_loaders(TRAIN_DATA_PATH, TEST_DATA_PATH,
val_split = 0.2, batch_size = BATCH_SIZE,
Expand Down
7 changes: 2 additions & 5 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Classification/Model_I/train.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -73,7 +73,7 @@ def fit_model(model, criterion, optimizer):
loss_dict = {'train_loss':[],'val_loss':[]}
acc_dict = {'train_accuracy':[],'val_accuracy':[]}

scheduler = CosineAnnealingWarmRestarts(optimizer,T_0 = 15, T_mult = 1,eta_min = 1e-6, verbose=True)
scheduler = CosineAnnealingWarmRestarts(optimizer, T_0=15, T_mult=1, eta_min=1e-6)

for epoch in range(EPOCHS):
print(f"Epoch {epoch+1}/{EPOCHS}:")
Expand All @@ -100,10 +100,7 @@ def fit_model(model, criterion, optimizer):
model = EffNetB2_backbone_model().to(device)

if LOAD_PRETRAINED_MODEL:
if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

criterion = nn.CrossEntropyLoss()
optimizer = optim.Adam(model.parameters(),lr = LEARNING_RATE)
Expand Down
5 changes: 1 addition & 4 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Classification/Model_II/test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -64,10 +64,7 @@ def flatten_list(x):
criterion = nn.CrossEntropyLoss()
model = EffNetB1_backbone_model().to(device)

if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

_, _, test_loader, class_map = create_data_loaders(TRAIN_DATA_PATH, TEST_DATA_PATH,
val_split = 0.2, batch_size = BATCH_SIZE,
Expand Down
7 changes: 2 additions & 5 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Classification/Model_II/train.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -75,7 +75,7 @@ def fit_model(model, criterion, optimizer):
loss_dict = {'train_loss':[],'val_loss':[]}
acc_dict = {'train_accuracy':[],'val_accuracy':[]}

scheduler = CosineAnnealingWarmRestarts(optimizer,T_0 = 15, T_mult = 1,eta_min = 1e-6, verbose=True)
scheduler = CosineAnnealingWarmRestarts(optimizer, T_0=15, T_mult=1, eta_min=1e-6)

# scheduler = ReduceLROnPlateau(optimizer, 'min',patience=2,factor=0.3,verbose=True)

Expand Down Expand Up @@ -105,10 +105,7 @@ def fit_model(model, criterion, optimizer):
model = EffNetB1_backbone_model().to(device)

if LOAD_PRETRAINED_MODEL:
if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

criterion = nn.CrossEntropyLoss()
optimizer = optim.Adam(model.parameters(),lr = LEARNING_RATE)
Expand Down
5 changes: 1 addition & 4 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Classification/Model_III/test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -64,10 +64,7 @@ def flatten_list(x):
criterion = nn.CrossEntropyLoss()
model = EffNetB1_backbone_model().to(device)

if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

_, _, test_loader, class_map = create_data_loaders(TRAIN_DATA_PATH, TEST_DATA_PATH,
val_split = 0.2, batch_size = BATCH_SIZE,
Expand Down
7 changes: 2 additions & 5 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Classification/Model_III/train.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -75,7 +75,7 @@ def fit_model(model, criterion, optimizer):
loss_dict = {'train_loss':[],'val_loss':[]}
acc_dict = {'train_accuracy':[],'val_accuracy':[]}

scheduler = CosineAnnealingWarmRestarts(optimizer,T_0 = 15, T_mult = 1,eta_min = 1e-6, verbose=True)
scheduler = CosineAnnealingWarmRestarts(optimizer, T_0=15, T_mult=1, eta_min=1e-6)

# scheduler = ReduceLROnPlateau(optimizer, 'min',patience=2,factor=0.3,verbose=True)

Expand Down Expand Up @@ -105,10 +105,7 @@ def fit_model(model, criterion, optimizer):
model = EffNetB1_backbone_model().to(device)

if LOAD_PRETRAINED_MODEL:
if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

criterion = nn.CrossEntropyLoss()
optimizer = optim.Adam(model.parameters(),lr = LEARNING_RATE)
Expand Down
5 changes: 1 addition & 4 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Regression/Model_I/test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,10 +20,7 @@ def flatten_list(x):
set_seed(7)
criterion = nn.HuberLoss()
model = Regressor().to(device)
if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

train_loader, val_loader, test_loader = create_dataloaders(
TRAIN_DATA_PATH, TEST_DATA_PATH,
Expand Down
8 changes: 2 additions & 6 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Regression/Model_I/train.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,3 @@
import gc
import numpy as np
from tqdm import tqdm
import matplotlib.pyplot as plt
Expand Down Expand Up @@ -99,7 +98,7 @@ def flatten_list(x):
def fit_model(model, checkpoint_path):

optimizer = Ranger(model.parameters(), lr = LEARNING_RATE)
scheduler = optim.lr_scheduler.OneCycleLR(optimizer, LEARNING_RATE, epochs = EPOCHS, steps_per_epoch = len(train_loader), verbose = False)
scheduler = optim.lr_scheduler.OneCycleLR(optimizer, LEARNING_RATE, epochs=EPOCHS, steps_per_epoch=len(train_loader))
criterion = nn.HuberLoss()

loss_dict = {'train_loss':[],'val_loss':[]}
Expand Down Expand Up @@ -140,10 +139,7 @@ def fit_model(model, checkpoint_path):
model = Regressor().to(device)

if LOAD_PRETRAINED_MODEL:
if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

model, loss_dict = fit_model(model, checkpoint_path = MODEL_PATH)

Expand Down
5 changes: 1 addition & 4 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Regression/Model_II/test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,10 +20,7 @@ def flatten_list(x):
set_seed(7)
criterion = nn.MSELoss()
model = Regressor().to(device)
if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

train_loader, val_loader, test_loader = create_dataloaders(
TRAIN_DATA_PATH, TEST_DATA_PATH,
Expand Down
8 changes: 2 additions & 6 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Regression/Model_II/train.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,3 @@
import gc
import numpy as np
from tqdm import tqdm
import matplotlib.pyplot as plt
Expand Down Expand Up @@ -99,7 +98,7 @@ def flatten_list(x):
def fit_model(model, checkpoint_path):

optimizer = Ranger(model.parameters(), lr = LEARNING_RATE)
scheduler = optim.lr_scheduler.OneCycleLR(optimizer, LEARNING_RATE, epochs = EPOCHS, steps_per_epoch = len(train_loader), verbose = False)
scheduler = optim.lr_scheduler.OneCycleLR(optimizer, LEARNING_RATE, epochs=EPOCHS, steps_per_epoch=len(train_loader))
criterion = nn.MSELoss()

loss_dict = {'train_loss':[],'val_loss':[]}
Expand Down Expand Up @@ -140,10 +139,7 @@ def fit_model(model, checkpoint_path):
model = Regressor().to(device)

if LOAD_PRETRAINED_MODEL:
if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

model, loss_dict = fit_model(model, checkpoint_path = MODEL_PATH)

Expand Down
5 changes: 1 addition & 4 deletions Updating_the_DeepLense_Pipeline__Saranga_K_Mahanta/Regression/Model_III/test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,10 +20,7 @@ def flatten_list(x):
set_seed(7)
criterion = nn.MSELoss()
model = Regressor().to(device)
if device != 'cpu':
model.load_state_dict(torch.load(MODEL_PATH))
else:
model.load_state_dict(torch.load(MODEL_PATH, map_location = torch.device('cpu')))
model.load_state_dict(torch.load(MODEL_PATH, map_location=device, weights_only=True))

train_loader, val_loader, test_loader = create_dataloaders(
TRAIN_DATA_PATH, TEST_DATA_PATH,
Expand Down
Loading