def do_update(self, line):
"""Update a given instance of a model"""
try:
tokens = split(line)
except ValueError:
return None
if len(tokens) < 1:
print("** class name missing **")
else:
objects = models.storage.all()
cls = models.getmodel(tokens[0])
if cls is None:
print("** class doesn't exist **")
elif len(tokens) < 2:
print("** instance id missing **")
elif ".".join(tokens[:2]) not in objects:
print("** no instance found **")
elif len(tokens) < 3:
print("** attribute name missing **")
elif len(tokens) < 4:
print("** value missing **")
else:
obj = objects[".".join(tokens[:2])]
for key, value in zip(tokens[2::2], tokens[3::2]):
try:
setattr(obj, key, int(value))
except ValueError:
try:
setattr(obj, key, float(value))
except ValueError:
try:
setattr(obj, key, str(value))
except ValueError:
pass
obj.save()
def reload(self):
"""
Load the dictionary of saved model instances from the filesystem
"""
try:
with open(self.__class__.__file_path, "r") as ifile:
objects = load(ifile)
for key, val in objects.items():
cls = models.getmodel(key.split(".")[0])
if cls:
self.__class__.__objects[key] = cls(**val)
except FileNotFoundError:
pass
def do_create(self, line):
"""Instantiate a given model"""
try:
tokens = split(line)
except ValueError:
return None
if len(tokens) < 1:
print("** class name missing **")
else:
cls = models.getmodel(tokens[0])
if cls is None:
print("** class doesn't exist **")
else:
instance = cls()
models.storage.save()
print(instance.id)
def do_all(self, line):
"""Show all instances of a given model or if unspecified, all models"""
try:
tokens = split(line)
except ValueError:
return
if len(tokens) < 1:
objects = models.storage.all()
print([str(obj) for obj in objects.values()])
else:
cls = models.getmodel(tokens[0])
if cls is None:
print("** class doesn't exist **")
return
objects = models.storage.all()
print([str(obj) for obj in objects.values() if type(obj) is cls])
def do_count(self, line):
"""Count the instances of a given model"""
try:
tokens = split(line)
except ValueError:
return None
objects = models.storage.all()
if len(tokens) < 1:
print("** class name missing **")
else:
cls = models.getmodel(tokens[0])
if cls is None:
print("** class doesn't exist **")
else:
matches = 0
for obj in objects.values():
if type(obj) is cls:
matches += 1
print(matches)
def do_show(self, line):
"""Show a given instance of a model"""
try:
tokens = split(line)
except ValueError:
return None
if len(tokens) < 1:
print("** class name missing **")
else:
objects = models.storage.all()
cls = models.getmodel(tokens[0])
if cls is None:
print("** class doesn't exist **")
elif len(tokens) < 2:
print("** instance id missing **")
elif ".".join(tokens[:2]) not in objects:
print("** no instance found **")
else:
print(objects[".".join(tokens[:2])])
def do_destroy(self, line):
"""Delete a given instance of a model"""
try:
tokens = split(line)
except ValueError:
return
if len(tokens) < 1:
print("** class name missing **")
return
cls = models.getmodel(tokens[0])
if cls is None:
print("** class doesn't exist **")
return
if len(tokens) < 2:
print("** instance id missing **")
return
objects = models.storage.all()
key = ".".join(tokens[0:2])
if key not in objects:
print("** no instance found **")
return
del objects[key]
models.storage.save()
dataLoaderTrain = DataLoader(dataset=datasetTrain,
batch_size=trBatchSize,
shuffle=True,
num_workers=24,
pin_memory=True)
dataLoaderVal = DataLoader(dataset=datasetValid,
batch_size=trBatchSize,
shuffle=False,
num_workers=24,
pin_memory=True)
dataLoaderTest = DataLoader(dataset=datasetTest,
num_workers=24,
pin_memory=True)
model = mod.getmodel(modelName, nnClassCount)
model = torch.nn.DataParallel(model)
if action == "train":
# train the model
timestampTime = time.strftime("%H%M%S")
timestampDate = time.strftime("%d%m%Y")
timestampLaunch = timestampDate + '-' + timestampTime
batch, losst, losse = CheXpertTrainer.train(CheXpertTrainer,
model,
dataLoaderTrain,
dataLoaderVal,
nnClassCount,
trMaxEpoch,
timestampLaunch,
dataLoaderTrain = DataLoader(dataset=datasetTrain,
batch_size=trBatchSize,
shuffle=True,
num_workers=24,
pin_memory=True)
dataLoaderVal = DataLoader(dataset=datasetValid,
batch_size=trBatchSize,
shuffle=False,
num_workers=24,
pin_memory=True)
dataLoaderTest = DataLoader(dataset=datasetTest,
num_workers=24,
pin_memory=True)
# initialize and load the model
model = mod.getmodel(modelName, nnClassCount)
model = torch.nn.DataParallel(model).cuda()
if action == "train":
# train the model
timestampTime = time.strftime("%H%M%S")
timestampDate = time.strftime("%d%m%Y")
timestampLaunch = timestampDate + '-' + timestampTime
batch, losst, losse = CheXpertTrainer.train(CheXpertTrainer,
model,
dataLoaderTrain,
dataLoaderVal,
nnClassCount,
trMaxEpoch,
timestampLaunch,
