def __init__(self, training, testing, model):
self.model = model
self.train_loader = dataset.loader(training)
self.loss = torch.nn.CrossEntropyLoss()
self.test_loader = dataset.test_loader(testing)
self.model.train()
self.prunner = FilterPrunner(self.model)
def __init__(self, train_path, test_path, model):
self.train_data_loader = dataset.loader(train_path)
self.test_data_loader = dataset.test_loader(test_path)
self.model = model
self.criterion = torch.nn.CrossEntropyLoss()
#self.prunner = FilterPr
self.model.train() #??
def __init__(self, train_path, test_path, model, optimizer):
self.train_data_loader = dataset.loader(train_path)
self.test_data_loader = dataset.test_loader(test_path)
self.optimizer = optimizer
self.model = model
self.criterion = torch.nn.CrossEntropyLoss()
self.model.train()
def __init__(self, train_path, test_path, model):
print(' PrunningFineTuner_AlexNet init')
self.train_data_loader = dataset.loader(train_path)
self.test_data_loader = dataset.test_loader(test_path)
self.model = model
self.criterion = torch.nn.CrossEntropyLoss()
print('PrunningFineTuner_AlexNet init filter prunner')
self.prunner = FilterPrunner(self.model)
def __init__(self, train_path, test_path, model):
self.train_data_loader = dataset.loader(train_path)
self.test_data_loader = dataset.test_loader(test_path)
#self.test_single_loader = dataset.test_loader(test_single_path) # I am adding this line...
self.model = model
self.criterion = torch.nn.CrossEntropyLoss()
self.prunner = FilterPrunner(self.model)
self.model.train()
def __init__(self, model, args):
#self.train_data_loader,self.class_dict = dataset.loader(args.train_path, num_workers=args.num_workers,no_crop=args.no_crop,grayscale=args.grayscale)
#self.test_data_loader = dataset.test_loader(args.test_path, num_workers=args.num_workers,no_crop=args.no_crop,grayscale=args.grayscale)
self.train_data_loader, self.class_dict = dataset.loader(
args.train_path, num_workers=args.num_workers)
self.test_data_loader = dataset.test_loader(
args.test_path, num_workers=args.num_workers)
self.model = model
self.criterion = torch.nn.CrossEntropyLoss()
self.model.train()
self.num_classes = args.num_classes
self.seed = args.seed
def __init__(self, model, args):
#self.train_data_loader,self.class_dict = dataset.loader(args.train_path, num_workers=args.num_workers,no_crop=args.no_crop,grayscale = args.grayscale)
#self.test_data_loader = dataset.test_loader(args.test_path, num_workers=args.num_workers,no_crop=args.no_crop,grayscale = args.grayscale)
self.train_data_loader, self.class_dict = dataset.loader(
args.train_path, num_workers=args.num_workers)
self.test_data_loader = dataset.test_loader(
args.test_path, num_workers=args.num_workers)
self.model = model
self.num_classes = self.model.out_channels
self.criterion = torch.nn.CrossEntropyLoss()
self.prunner = FilterPrunner(self.model)
self.start = args.start
self.indiv_acc = args.indiv_acc
self.model.train()
self.args = args
maxlen = int(maxlen)
folder = "/scratch/home/Projects/rnn_dropout/exps/"
filename = ("sa_DropoutLSTM_pW_%.2f_pU_%.2f_pDense_%.2f_pEmb_%.2f_reg_%f_batch_size_%d_cutoff_%d_epochs"
% (p_W, p_U, p_dense, p_emb, weight_decay, batch_size, maxlen))
print(filename)
# Global params:
nb_words = 20000
skip_top = 0
test_split = 0.2
init_seed = 0
global_seed = 0
# Load data:
print("Loading data...")
dataset = loader(init_seed, maxlen, nb_words, skip_top, test_split)
X_train, X_test, Y_train, Y_test = dataset.X_train, dataset.X_test, dataset.Y_train, dataset.Y_test
mean_y_train, std_y_train = dataset.mean_y_train, dataset.std_y_train
# Set seed:
np.random.seed(global_seed)
# Build model:
print('Build model...')
model = Sequential()
model.add(Embedding(nb_words + dataset.index_from, 128, W_regularizer=l2(weight_decay),
dropout=p_emb, input_length=maxlen, batch_input_shape=(batch_size, maxlen)))
model.add(LSTM(128, W_regularizer=l2(weight_decay), U_regularizer=l2(weight_decay),
b_regularizer=l2(weight_decay), dropout_W=p_W, dropout_U=p_U))
model.add(Dropout(p_dense))
model.add(Dense(1, W_regularizer=l2(weight_decay), b_regularizer=l2(weight_decay)))
preds = best_estimator.predict(X_test)
mse = mean_squared_error(y_test, preds)
return runtime, mse
def write_results(
results
): # Writes benchmarking results to a CSV file (pandas for this is overkill,
df = DataFrame(results) # but I'm lazy
df.columns = ['Model', 'Runtime', 'MSE']
df.to_csv("results.csv", index=False)
if __name__ == "__main__":
print("Loading data...")
X, y = loader()
# del X, y
print("Data loaded")
xgb = XGBRegressor(n_estimators=n_estimators)
dxgb = dXGBRegressor(n_estimators=n_estimators)
descriptions = ["XGBoost", "XGBoost (dask-ml)"]
results = []
for estimator, description in zip([xgb, dxgb], descriptions):
runtime, mse = benchmark(dxgb, param_grid, X, y)
print("Model: {}".format(description))
print("Runtime: {} s".format(runtime))
print("MSE: {}".format(mse))
results.append((description, runtime, mse))
model.eval()
pred = model(data, target)
tg += sum(torch.eq(torch.LongTensor(p), t).sum() for p, t in zip(pred, target))
tt += sum(map(len, data))
# Eval
model.eval()
with torch.no_grad():
eg, et = 0, 0
for data, target in eval_loader:
pred = model(data, target)
eg += sum(torch.eq(torch.LongTensor(p), t).sum() for p, t in zip(pred, target)).int()
et += sum(map(len, data))
logging('Done epoch {}/{} ({} batches) train accuracy {:.2f}, eval accuracy {:.2f} avg loss {:.5f}'.format(
epoch+1, epochs, (epoch+1)*train_loader.__len__(), float(tg) / tt, float(eg) / et, avg_loss))
def logging(message):
print('{} {}'.format(dtm.datetime.now().strftime("%Y-%m-%d %H:%M:%S.%f")[:-3], message))
if __name__ == "__main__":
torch.manual_seed(1)
device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
data = TagData('pos')
model = LSTMCRF(vocab_size=len(data.words), embed_dim=100, hidden_dim=100
, num_tags=len(data.tags), bidirectional=True, num_layers=2, device=device, dropout=0.5).to(device)
optimizer = Adam(model.parameters(), lr=0.01)
train_crf(model, 10, optimizer, loader(data, 'train', 1000), loader(data, 'dev', 10), device=device)
