Create app.py

app.py

@@ -0,0 +1,36 @@
+import torch
+import torch.nn as nn
+import numpy as np
+
+SEQUENCE_LENGTH = 10
+INPUT_SIZE = 1
+OUTPUT_SIZE = 1
+HIDDEN_UNITS = 128
+device = torch.device('cpu')
+
+class Seq2Seq(nn.Module):
+    def __init__(self, input_size, hidden_size, output_size, seq_len):
+        super(Seq2Seq, self).__init__()
+        self.seq_len = seq_len
+        self.hidden_size = hidden_size
+
+        self.encoder_lstm = nn.LSTM(input_size, hidden_size, batch_first=True)
+        self.decoder_lstm = nn.LSTM(hidden_size, hidden_size, batch_first=True)
+        self.decoder_linear = nn.Linear(hidden_size, output_size)
+
+    def forward(self, x):
+        _, (hidden, cell) = self.encoder_lstm(x)
+        
+        context_vector = hidden.permute(1, 0, 2)
+        decoder_input = context_vector.repeat(1, self.seq_len, 1)
+        
+        decoder_output, _ = self.decoder_lstm(decoder_input, (hidden, cell))
+        
+        prediction = self.decoder_linear(decoder_output)
+        
+        return prediction
+
+model_path = 'seq2seq_model_weights.pth' 
+model = Seq2Seq(INPUT_SIZE, HIDDEN_UNITS, OUTPUT_SIZE, SEQUENCE_LENGTH).to(device)
+model.load_state_dict(torch.load(model_path, map_location=device))
+model.eval()