def main():
if len(sys.argv) <= 2:
print(
"please specify a number of examples and a model name (e.g. models.baseline.random_handle)"
)
sys.exit(1)
eval_set_size = int(sys.argv[1])
module_name = sys.argv[2]
# splitting training data
print("splitting training data into", eval_set_size,
"(test) v. rest (train)")
data.load_train()
tweets = np.array(data.TRAIN)
np.random.seed(SEED)
np.random.shuffle(tweets)
test_tweets, train_tweets = tweets[:eval_set_size], tweets[eval_set_size:]
hyper_parameters = models.parse_hyper_parameters(sys.argv[3:])
model_class = importlib.import_module(module_name).Model
print("Model:", module_name, hyper_parameters)
print("Training...")
model = model_class(tqdm(train_tweets, dynamic_ncols=True),
**hyper_parameters)
print("Evaluating...")
accuracy, correct, tests = eval.evaluate(
model, tqdm(test_tweets, dynamic_ncols=True))
print(f"Label accuracy: {correct}/{tests} ({accuracy:%})")
def main():
if len(sys.argv) <= 1:
print(
"please specify a model name (e.g. models.baseline.random_handle)")
sys.exit(1)
module_name = sys.argv[1]
hyper_parameters = models.parse_hyper_parameters(sys.argv[2:])
model = models.load(module_name, hyper_parameters)
data.load_devel1000()
print("Model:", module_name, hyper_parameters)
accuracy, correct, tests = evaluate(
model, tqdm(data.DEVEL1000, dynamic_ncols=True))
print(f"Label accuracy: {correct}/{tests} ({accuracy:%})")
def main():
if len(sys.argv) <= 1:
print(
"please specify a model name (e.g. models.baseline.random_handle)")
sys.exit(1)
module_name = sys.argv[1]
hyper_parameters = models.parse_hyper_parameters(sys.argv[2:])
model_class = importlib.import_module(module_name).Model
print("Model:", module_name, hyper_parameters)
print("Training...")
model = model_class(tqdm(data.TRAIN, dynamic_ncols=True),
**hyper_parameters)
print("Evaluating...")
accuracy, correct, tests = eval.evaluate(
model, tqdm(data.DEVEL, dynamic_ncols=True))
print(f"Label accuracy: {correct}/{tests} ({accuracy:%})")
def main():
# interpret command line arguments
if len(sys.argv) <= 1:
print(
"please specify a model name (e.g. models.baseline.random_handle)")
sys.exit(1)
module_name = sys.argv[1]
hyper_parameters = models.parse_hyper_parameters(sys.argv[2:])
# training model
module = importlib.import_module(module_name)
print(
f"Training {module_name}.Model with hyperparameters {hyper_parameters}"
)
model = module.Model(tqdm(data.TRAIN, dynamic_ncols=True),
**hyper_parameters)
print("Training done!")
models.save(model, module_name, hyper_parameters)
def main():
if len(sys.argv) <= 1:
print(
"please specify a model name (e.g. models.baseline.random_handle)")
sys.exit(1)
module_name = sys.argv[1]
hyper_parameters = models.parse_hyper_parameters(sys.argv[2:])
model_class = importlib.import_module(module_name).Model
print("Model:", module_name, hyper_parameters)
print("Training on ALL of the training data...")
model = model_class(tqdm(data.ALL_TRAIN, dynamic_ncols=True),
**hyper_parameters)
print("Labelling unlabelled tweets...")
for tweet in tqdm(data.TEST, dynamic_ncols=True):
tweet.handle = model.predict(tweet)
print("Saving submission...")
current_time = time.strftime('%m-%d_%H-%M-%S')
model_name = models.model_name(module_name, hyper_parameters)
filename = f"../submissions/submission-{current_time}-{model_name}.csv"
data.export(filename, data.TEST)
print("Saved to:", filename)