Modified the code for frequentist experiment
This commit is contained in:
Eduardo Cueto
parent
a2300273a5
commit
0f8ee842a3
|
|
@ -1,19 +1,21 @@
|
|||
__pycache__/
|
||||
checkpoints/
|
||||
data/
|
||||
data_budget/
|
||||
experiment-power-draw/
|
||||
**/__pycache__/
|
||||
**/__init__.py
|
||||
**/**/__pycache__/
|
||||
**/**/__init__.py
|
||||
stp
|
||||
sav
|
||||
**/__pycache__/
|
||||
**/__init__.py
|
||||
checkpoints/
|
||||
__pycache__/
|
||||
data_budget/
|
||||
bayes_*
|
||||
times_*
|
||||
.vscode
|
||||
freq_*
|
||||
data/
|
||||
.idea
|
||||
*.pkl
|
||||
*.txt
|
||||
stp
|
||||
sav
|
||||
bay
|
||||
frq
|
||||
sav
|
||||
|
|
|
|||
|
|
@ -8,20 +8,26 @@ import metrics
|
|||
import argparse
|
||||
import numpy as np
|
||||
import torch.nn as nn
|
||||
import amd_sample_draw
|
||||
from datetime import datetime
|
||||
import config_frequentist as cfg
|
||||
from torch.optim import Adam, lr_scheduler
|
||||
from gpu_power_func import total_watt_consumed
|
||||
from models.NonBayesianModels.LeNet import LeNet
|
||||
from models.NonBayesianModels.AlexNet import AlexNet
|
||||
from stopping_crit import earlyStopping, energyBound, accuracyBound
|
||||
from models.NonBayesianModels.ThreeConvThreeFC import ThreeConvThreeFC
|
||||
|
||||
with (open("configuration.pkl", "rb")) as file:
|
||||
while True:
|
||||
try:
|
||||
cfg = pickle.load(file)
|
||||
except EOFError:
|
||||
break
|
||||
|
||||
# CUDA settings
|
||||
device = torch.device("cuda:1" if torch.cuda.is_available() else "cpu")
|
||||
|
||||
|
||||
def getModel(net_type, inputs, outputs,wide=cfg.wide):
|
||||
def getModel(net_type, inputs, outputs,wide=cfg["model"]["size"]):
|
||||
if (net_type == 'lenet'):
|
||||
return LeNet(outputs, inputs,wide)
|
||||
elif (net_type == 'alexnet'):
|
||||
|
|
@ -64,11 +70,11 @@ def validate_model(net, criterion, valid_loader):
|
|||
def run(dataset, net_type):
|
||||
|
||||
# Hyper Parameter settings
|
||||
n_epochs = cfg.n_epochs
|
||||
lr = cfg.lr
|
||||
num_workers = cfg.num_workers
|
||||
valid_size = cfg.valid_size
|
||||
batch_size = cfg.batch_size
|
||||
n_epochs = cfg["model"]["n_epochs"]
|
||||
lr = cfg["model"]["lr"]
|
||||
num_workers = cfg["model"]["num_workers"]
|
||||
valid_size = cfg["model"]["valid_size"]
|
||||
batch_size = cfg["model"]["batch_size"]
|
||||
|
||||
trainset, testset, inputs, outputs = data.getDataset(dataset)
|
||||
train_loader, valid_loader, test_loader = data.getDataloader(
|
||||
|
|
@ -76,15 +82,13 @@ def run(dataset, net_type):
|
|||
net = getModel(net_type, inputs, outputs).to(device)
|
||||
|
||||
ckpt_dir = f'checkpoints/{dataset}/frequentist'
|
||||
ckpt_name = f'checkpoints/{dataset}/frequentist/model_{net_type}_{cfg.wide}.pt'
|
||||
ckpt_name = f'checkpoints/{dataset}/frequentist/model_{net_type}_{cfg["model"]["size"]}.pt'
|
||||
|
||||
if not os.path.exists(ckpt_dir):
|
||||
os.makedirs(ckpt_dir, exist_ok=True)
|
||||
|
||||
with open("stp", "r") as file:
|
||||
stp = int(file.read())
|
||||
with open("sav", "r") as file:
|
||||
sav = int(file.read())
|
||||
stp = cfg["stopping_crit"]
|
||||
sav = cfg["save"]
|
||||
|
||||
criterion = nn.CrossEntropyLoss()
|
||||
optimizer = Adam(net.parameters(), lr=lr)
|
||||
|
|
@ -108,25 +112,25 @@ def run(dataset, net_type):
|
|||
|
||||
if stp == 2:
|
||||
#print('Using early stopping')
|
||||
if earlyStopping(early_stop,valid_acc,epoch,cfg.sens) == 1:
|
||||
if earlyStopping(early_stop,valid_acc,epoch,cfg["model"]["sens"]) == 1:
|
||||
break
|
||||
elif stp == 3:
|
||||
#print('Using energy bound')
|
||||
if energyBound(cfg.energy_thrs) == 1:
|
||||
if energyBound(cfg["model"]["energy_thrs"]) == 1:
|
||||
break
|
||||
elif stp == 4:
|
||||
#print('Using accuracy bound')
|
||||
if accuracyBound(train_acc,cfg.acc_thrs) == 1:
|
||||
if accuracyBound(train_acc,cfg["model"]["acc_thrs"]) == 1:
|
||||
break
|
||||
else:
|
||||
print('Training for {} epochs'.format(cfg.n_epochs))
|
||||
print('Training for {} epochs'.format(cfg["model"]["n_epochs"]))
|
||||
|
||||
if sav == 1:
|
||||
# save model when finished
|
||||
if epoch == n_epochs:
|
||||
torch.save(net.state_dict(), ckpt_name)
|
||||
|
||||
with open("freq_exp_data_"+str(cfg.wide)+".pkl", 'wb') as f:
|
||||
with open("freq_exp_data_"+str(cfg["model"]["size"])+".pkl", 'wb') as f:
|
||||
pickle.dump(train_data, f)
|
||||
|
||||
|
||||
|
|
@ -134,11 +138,7 @@ if __name__ == '__main__':
|
|||
now = datetime.now()
|
||||
current_time = now.strftime("%H:%M:%S")
|
||||
print("Initial Time =", current_time)
|
||||
parser = argparse.ArgumentParser(description = "PyTorch Frequentist Model Training")
|
||||
parser.add_argument('--net_type', default='lenet', type=str, help='model')
|
||||
parser.add_argument('--dataset', default='MNIST', type=str, help='dataset = [MNIST/CIFAR10/CIFAR100]')
|
||||
args = parser.parse_args()
|
||||
run(args.dataset, args.net_type)
|
||||
run(cfg["model"]["data"], cfg["model"]["net_type"])
|
||||
now = datetime.now()
|
||||
current_time = now.strftime("%H:%M:%S")
|
||||
print("Final Time =", current_time)
|
||||
|
|
|
|||
|
|
@ -60,34 +60,20 @@ cfg["model"]["size"] = wide
|
|||
cfg["data"] = args["dataset"]
|
||||
cfg["model"]["net_type"] = args["net_type"]
|
||||
|
||||
#with open("tmp", "w") as file:
|
||||
# file.write(str(wide))
|
||||
|
||||
if args['EarlyStopping']:
|
||||
cfg["stopping_crit"] = 2
|
||||
#with open("stp", "w") as file:
|
||||
# file.write('2')
|
||||
elif args['EnergyBound']:
|
||||
cfg["stopping_crit"] = 3
|
||||
#with open("stp", "w") as file:
|
||||
# file.write('3')
|
||||
elif args['AccuracyBound']:
|
||||
cfg["stopping_crit"] = 4
|
||||
#with open("stp", "w") as file:
|
||||
# file.write('4')
|
||||
else:
|
||||
cfg["stopping_crit"] = 1
|
||||
#with open("stp", "w") as file:
|
||||
# file.write('1')
|
||||
|
||||
if args['Save']:
|
||||
cfg["save"] = 1
|
||||
#with open("sav", "w") as file:
|
||||
# file.write('1')
|
||||
else:
|
||||
cfg["save"] = 0
|
||||
#with open("sav", "w") as file:
|
||||
# file.write('0')
|
||||
|
||||
|
||||
cfg["pickle_path"] = "{}_wattdata_{}.pkl".format(cfg["model"]["type"],cfg["model"]["size"])
|
||||
|
|
@ -96,7 +82,7 @@ with open("configuration.pkl", "wb") as f:
|
|||
pickle.dump(cfg, f)
|
||||
|
||||
#print(args)
|
||||
print(cfg)
|
||||
#print(cfg)
|
||||
|
||||
sleep(3)
|
||||
|
||||
|
|
@ -105,18 +91,10 @@ if cmd[1] == "main_frequentist.py":
|
|||
cmd2 = ["./cpu_watt.sh", "freq_{}_cpu_watts".format(wide)]
|
||||
cmd3 = ["./mem_free.sh", "freq_{}_ram_use".format(wide)]
|
||||
cmd4 = ["./radeontop.sh", "freq_{}_flop_app".format(wide)]
|
||||
#with open("frq", "w") as file:
|
||||
# file.write(str(1))
|
||||
#with open("bay", "w") as file:
|
||||
# file.write(str(0))
|
||||
elif cmd[1] == "main_bayesian.py":
|
||||
cmd2 = ["./cpu_watt.sh", "bayes_{}_cpu_watts".format(wide)]
|
||||
cmd3 = ["./mem_free.sh", "bayes_{}_ram_use".format(wide)]
|
||||
cmd4 = ["./radeontop.sh", "bayes_{}_flop_app".format(wide)]
|
||||
#with open("bay", "w") as file:
|
||||
# file.write(str(1))
|
||||
#with open("frq", "w") as file:
|
||||
# file.write(str(0))
|
||||
|
||||
|
||||
path = sub.check_output(['pwd'])
|
||||
|
|
@ -126,18 +104,9 @@ path = path.replace('\n', '')
|
|||
startWattCounter = 'python ' + path + '/amd_sample_draw.py'
|
||||
|
||||
|
||||
#test = startNODE.split()
|
||||
#test.append(pythonEnd)
|
||||
#test.append(pythonEnd2)
|
||||
|
||||
#startNODE = test
|
||||
|
||||
#print(startNODE)
|
||||
#print(startWattCounter)
|
||||
|
||||
p1 = sub.Popen(cmd)
|
||||
#p2 = sub.Popen(startWattCounter.split(),stdin=sub.PIPE,stdout=sub.PIPE, stderr=sub.PIPE)
|
||||
p2 = sub.Popen(startWattCounter.split())
|
||||
p2 = sub.Popen(startWattCounter.split(),stdin=sub.PIPE,stdout=sub.PIPE, stderr=sub.PIPE)
|
||||
p3 = sub.Popen(cmd2,stdin=sub.PIPE,stdout=sub.PIPE, stderr=sub.PIPE)
|
||||
p4 = sub.Popen(cmd3,stdin=sub.PIPE,stdout=sub.PIPE, stderr=sub.PIPE)
|
||||
p5 = sub.Popen(cmd4,stdin=sub.PIPE,stdout=sub.PIPE, stderr=sub.PIPE)
|
||||
|
|
|
|||
Loading…
Reference in New Issue