def log_trial_end(self, trial, failed):
e = CNVRGExperiment(self._cnvrg_experiments[trial.trial_id])
e.log("===== Logging Artifacts =====")
from os import listdir
files_list = [
os.path.join(trial.logdir, p) for p in os.listdir(trial.logdir)
]
e.log_artifacts(files_list)
e.finish(exit_status=int(failed))
# Name generator
names = get_images_name(os.path.join(args.evalf, "images"))
model.eval()
with torch.no_grad():
for data, target in eval_loader:
data, target = data.to(device), target.to(device)
output = model(data)
label = output.argmax(dim=1, keepdim=True).item()
print ("Images: " + next(names) + ", Classified as: " + str(label))
# Train?
if args.train:
# Train + Test per epoch
for epoch in range(1, args.epochs + 1):
train(args, model, device, train_loader, optimizer, epoch)
accuracy = test(args, model, device, test_loader, epoch)
# Do checkpointing - Is saved in outf
outfile = '%s/mnist_convnet_model_epoch_%d.pth' % (args.outf, args.epochs)
torch.save(model.state_dict(), outfile)
e.log_artifacts([outfile])
# Evaluate?
if args.evaluate:
accuracy = test_image()
e.log_param("accuracy", accuracy)
e.finish()