Modified the code for frequentist experiment

.gitignore

@@ -1,19 +1,21 @@
-__pycache__/
-checkpoints/
-data/
-data_budget/
 experiment-power-draw/
-**/__pycache__/
-**/__init__.py
 **/**/__pycache__/
 **/**/__init__.py
-stp
+**/__pycache__/
-sav
+**/__init__.py
+checkpoints/
+__pycache__/
+data_budget/
 bayes_*
 times_*
+.vscode
 freq_*
+data/
+.idea
 *.pkl
 *.txt
+stp
+sav
 bay
 frq
 sav

main_frequentist.py

@@ -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
+    n_epochs = cfg["model"]["n_epochs"]
-    lr = cfg.lr
+    lr = cfg["model"]["lr"]
-    num_workers = cfg.num_workers
+    num_workers = cfg["model"]["num_workers"]
-    valid_size = cfg.valid_size
+    valid_size = cfg["model"]["valid_size"]
-    batch_size = cfg.batch_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 = cfg["stopping_crit"]
-        stp = int(file.read())
+    sav = cfg["save"]
-    with open("sav", "r") as file:
-        sav = int(file.read())
 
     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")
+    run(cfg["model"]["data"], cfg["model"]["net_type"])
-    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)
     now = datetime.now()
     current_time = now.strftime("%H:%M:%S")
     print("Final Time =", current_time)

run_service.py

@@ -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(),stdin=sub.PIPE,stdout=sub.PIPE, stderr=sub.PIPE)
-p2 = sub.Popen(startWattCounter.split())
 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)