preparing for batch job rnn 73 days

create_submission.py

@@ -37,7 +37,7 @@ def create_submission(model, transform, device):
     # is not using the temporal context, this is here arbitrarily chosen
     # However, note that it must be a divisor of the total number of days
     # in the 2017 year , either 1, 5, 73 or 365
-    num_days = 365
+    num_days = 73
     num_workers = 7
 
     use_cuda = torch.cuda.is_available()

job.py

@@ -75,7 +75,9 @@ parser.add_argument("--model_name",
 # Ensure the log directory exists
 os.system("mkdir -p logslurms")
 
+args = parser.parse_args()
+
 time_wall = {"no_limit": "48:00:00","hour" : "1:00:00", "half" : "0:00:00", "quarter" : "0:00:15"}
 
 # Launch the batch jobs
-submit_job(makejob(commit_id, "cnn", 1, args.user, time_wall[args.time_wall]))
\ No newline at end of file
+submit_job(makejob(commit_id, args.model_name, 1, args.user, time_wall[args.time_wall]))
\ No newline at end of file

main.py

@@ -33,18 +33,19 @@ if __name__ == "__main__":
 
     parser.add_argument(
     "--rootDir",
+    default=None,
     help="Directory in which the log files will be stored"
     )
 
     args = parser.parse_args()
 
+    config_file = open("config.yml")
+    cfg = yaml.load(config_file)
 
-    rootDir = cfg["LogDir"] if eval(args.rootDir) != None else args.rootDir
+    rootDir = args.rootDir if args.rootDir != None else cfg["LogDir"]
 
     logging.basicConfig(filename= rootDir + 'main_unit_test.log', level=logging.INFO)
     
-    config_file = open("config.yml")
-    cfg = yaml.load(config_file)
 
     use_cuda = torch.cuda.is_available()
     trainpath           = cfg["Dataset"]["_DEFAULT_TRAIN_FILEPATH"]
@@ -71,7 +72,7 @@ if __name__ == "__main__":
         MAX    = eval(cfg["ApproximativeMaxi"])
         MIN    = eval(cfg["ApproximativeMini"])
     else :
-        MEAN, STD, MAX, MIN = get_stats_train_dataset(trainpath,
+        MEAN, STD, MAX, MIN = dataloader.get_stats_train_dataset(trainpath,
         num_days,
         batch_size,
         num_workers,

model.py

@@ -3,6 +3,7 @@ import torch
 import torch.nn as nn
 import torch.nn.functional as F
 from torch.autograd import Function
+from torch.autograd import Variable
 
 import torch.nn as nn
 
@@ -35,6 +36,32 @@ class LinearRegression(nn.Module):
     def forward(self, x):
         return self.regressor(x) 
 
+# Recurrent Neural Networks
+class RNN(nn.Module):
+    def __init__(self, cfg, input_size):
+        super(RNN, self).__init__()
+        self.hidden_size = cfg["RNN"]["HiddenSize"]
+        self.num_layers = cfg["RNN"]["NumLayers"]
+        # RNN
+        self.rnn = nn.RNN(input_size, self.hidden_size, self.num_layers, batch_first=True, nonlinearity='relu')
+        
+        self.fc = nn.Linear(self.hidden_size, 1)
+    
+    def forward(self, x):
+        use_cuda = torch.cuda.is_available()
+        if use_cuda :
+            device = torch.device('cuda')
+        else :
+            device = toch.device('cpu')
+        # Initialize hidden state with zeros
+        h0 = torch.zeros(self.num_layers, x.size(0), self.hidden_size).to(device)
+            
+        # One time step
+        out, hn = self.rnn(x, h0)
+        out = self.fc(out)
+        print(out.shape)
+        return out
+
 # Bidirectional LSTM
 class BidirectionalLSTM(nn.Module):
     def __init__(self, cfg, input_size):

train.py

@@ -31,6 +31,7 @@ def train(args, model, loader, f_loss, optimizer, device, log_interval = 100):
 
         # Compute the forward pass through the network up to the loss
         outputs = model(inputs)
+        print(targets.shape)
         loss = f_loss(outputs, targets)
 
         # Backward and optimize

wandb/debug-internal.log

-run-20230130_164435-f5w8sswr/logs/debug-internal.log
\ No newline at end of file
+run-20230131_200825-ytr4zy6r/logs/debug-internal.log
\ No newline at end of file

wandb/debug.log

-run-20230130_164435-f5w8sswr/logs/debug.log
\ No newline at end of file
+run-20230131_200825-ytr4zy6r/logs/debug.log
\ No newline at end of file

wandb/latest-run

-run-20230130_164435-f5w8sswr
\ No newline at end of file
+run-20230131_200825-ytr4zy6r
\ No newline at end of file