def test_fit_swa_cuda(self):
for model_name in supported_tv_models:
model = cnn.create_cnn(model_name, 10, pretrained=None)
opt = torch.optim.Adam(model.parameters(), lr=1e-3)
loss = nn.CrossEntropyLoss()
scheduler = torch.optim.lr_scheduler.CosineAnnealingLR(opt,
T_max=300)
swa_scheduler = SWALR(opt,
anneal_strategy="linear",
anneal_epochs=20,
swa_lr=0.05)
swa_start = 2
history = cnn.fit(model,
3,
train_loader,
val_loader,
loss,
device="cpu",
optimizer=opt,
scheduler=scheduler,
num_batches=10,
swa_start=swa_start,
swa_scheduler=swa_scheduler)
self.assertIsInstance(history, Dict)
exp_keys = ("train", "val")
for exp_k in exp_keys:
self.assertTrue(exp_k in history.keys())
exp_keys2 = ("top1_acc", "top5_acc", "loss")
for exp_k2 in exp_keys2:
self.assertTrue(exp_k2 in history["train"].keys())
self.assertTrue(exp_k2 in history["val"].keys())
def test_fit_cuda(self):
for model_name in supported_tv_models:
model = cnn.create_cnn(model_name, 10, pretrained=None)
opt = torch.optim.Adam(model.parameters(), lr=1e-3)
loss = nn.CrossEntropyLoss()
# scheduler = torch.optim.lr_scheduler.CyclicLR(optimizer=opt, base_lr=1e-4, max_lr=1e-3, mode="min")
history = cnn.fit(model,
1,
train_loader,
val_loader,
loss,
device="cuda",
optimizer=opt,
num_batches=10,
fp16=True)
self.assertIsInstance(history, Dict)
exp_keys = ("train", "val")
for exp_k in exp_keys:
self.assertTrue(exp_k in history.keys())
exp_keys2 = ("top1_acc", "top5_acc", "loss")
for exp_k2 in exp_keys2:
self.assertTrue(exp_k2 in history["train"].keys())
self.assertTrue(exp_k2 in history["val"].keys())
def test_val_sanity_fit(self):
for model_name in supported_tv_models:
model = cnn.create_cnn(model_name, 10, pretrained=None)
model = model.cpu()
loss = nn.CrossEntropyLoss()
res = cnn.val_sanity_fit(model,
val_loader,
loss,
"cpu",
num_batches=10)
self.assertTrue(res)
def test_infer(self):
# Infer over an image
image = Image.open("tests/assets/grace_hopper_517x606.jpg")
tensor = im2tensor(image)
self.assertEqual(tensor.ndim, 4)
for model_name in supported_tv_models:
model = cnn.create_cnn(model_name, 10, pretrained=None)
model = model.cpu()
model = model.eval()
out = model(tensor)
self.assertEqual(out.shape[1], 10)
self.assertEqual(out.ndim, 2)
def test_train(self):
# Read Image using PIL Here
# Do forward over image
image = Image.open("tests/assets/grace_hopper_517x606.jpg")
tensor = im2tensor(image)
self.assertEqual(tensor.ndim, 4)
for model_name in supported_tv_models:
model = cnn.create_cnn(model_name, 10, pretrained=None)
model = model.cpu()
out = model(tensor)
self.assertEqual(out.shape[1], 10)
self.assertEqual(out.ndim, 2)
def test_sanity_fit_cuda(self):
for model_name in supported_tv_models:
model = cnn.create_cnn(model_name, 10, pretrained=None)
loss = nn.CrossEntropyLoss()
res = cnn.sanity_fit(model,
train_loader,
val_loader,
loss,
"cuda",
num_batches=10,
fp16=True)
self.assertTrue(res)
def test_train_sanity_fit_cuda(self):
for model_name in supported_tv_models:
model = cnn.create_cnn(model_name, 10, pretrained=None)
opt = torch.optim.Adam(model.parameters(), lr=1e-3)
loss = nn.CrossEntropyLoss()
res = cnn.train_sanity_fit(model,
train_loader,
loss,
"cuda",
num_batches=10,
fp16=True)
self.assertTrue(res)
def test_val_step(self):
for model_name in supported_tv_models:
model = cnn.create_cnn(model_name, 10, pretrained=None)
model = model.cpu()
loss = nn.CrossEntropyLoss()
val_metrics = cnn.val_step(model,
val_loader,
loss,
"cpu",
num_batches=10)
self.assertIsInstance(val_metrics, Dict)
exp_keys = ("loss", "top1", "top5")
for exp_k in exp_keys:
self.assertTrue(exp_k in val_metrics.keys())
def test_csv_single_label_dataset(self):
complete_dataset = CSVSingleLabelDataset(df, data_dir, "Image",
"Label", tfms, "png")
self.assertTrue(complete_dataset[0])
train_loader = torch.utils.data.DataLoader(complete_dataset,
num_workers=1)
model = cnn.create_cnn("resnet18", 2, pretrained=None)
opt = torch.optim.Adam(model.parameters(), lr=1e-3)
loss = nn.CrossEntropyLoss()
res = cnn.train_sanity_fit(model,
train_loader,
loss,
"cpu",
num_batches=1)
self.assertTrue(res)
def test_train_step_cuda(self):
for model_name in supported_tv_models:
model = cnn.create_cnn(model_name, 10, pretrained=None)
opt = torch.optim.Adam(model.parameters(), lr=1e-3)
loss = nn.CrossEntropyLoss()
# scheduler = torch.optim.lr_scheduler.CyclicLR(optimizer=opt, base_lr=1e-4, max_lr=1e-3, mode="min")
scaler = amp.GradScaler()
train_metrics = cnn.train_step(model,
train_loader,
loss,
"cuda",
opt,
num_batches=10,
scaler=scaler)
self.assertIsInstance(train_metrics, Dict)
exp_keys = ("loss", "top1", "top5")
for exp_k in exp_keys:
self.assertTrue(exp_k in train_metrics.keys())
def test_create_cnn(self):
for model_name in supported_tv_models:
model = cnn.create_cnn(model_name, 10, pretrained=None)
self.assertTrue(isinstance(model, nn.Module))
