Fix mamba git conflicts

code/mamba/mamba.py

@@ -11,7 +11,6 @@ def __init__(self,
         dropout: float = 0.2,
         layer_norm_eps: float = 1e-5,
     ) -> None:
-<<<<<<< HEAD
         """ Mamba block
         
         Args:   
@@ -21,18 +20,6 @@ def __init__(self,
             expand (int): the expansion factor.
             dropout (float): the dropout rate.
             layer_norm_eps (float): the layer normalization epsilon.
-=======
-        """ Mamba block as described in the paper.
-
-        Args: 
-            d_model: The input and output feature dimension.
-            d_state: The state dimension.
-            d_conv: The convolutional dimension.
-            expand: The number of dimensions to expand.
-            dropout: The dropout rate.
-            layer_norm_eps: The epsilon value for layer normalization.
-            weight_decay: The L2 regularization
->>>>>>> 1e5bf639793de07acee6aaa4da95df9566f29c94
         """
         super().__init__()
         self.d_model = d_model
@@ -54,7 +41,6 @@ def __init__(self,
         self.us_proj = nn.Linear(d_state, d_model)
 
 
-<<<<<<< HEAD
     def forward(self, x: torch.Tensor) -> torch.Tensor:
         """ Forward pass
         
@@ -63,18 +49,6 @@ def forward(self, x: torch.Tensor) -> torch.Tensor:
 
         Returns: 
             torch.Tensor: the output tensor.
-=======
-    def forward(self, 
-        x: torch.Tensor
-    ) -> torch.Tensor:
-        """ Forward pass of the model.
-
-        Args:
-            x: The input tensor of shape (B, L, D).
-
-        Returns: 
-            The output tensor of shape (B, L, D). 
->>>>>>> 1e5bf639793de07acee6aaa4da95df9566f29c94
         """
         B, L, D = x.shape
 
@@ -107,6 +81,7 @@ def forward(self,
 
         return x
 
+
 class Mamba(nn.Module):
     def __init__(self, 
         d_model: int, 
@@ -119,7 +94,6 @@ def __init__(self,
         layer_norm_eps: float = 1e-5,
         spectral_norm: bool = True
     ):
-<<<<<<< HEAD
         """ Mamba model
         
         Args: 
@@ -135,24 +109,6 @@ def __init__(self,
 
         References: 
             1. Gu, A., & Dao, T. (2023). 
-=======
-        """ Mamba model as described in the paper.
-
-        Args: 
-            d_model: The input and output feature dimension.
-            d_state: The state dimension.
-            d_conv: The convolutional dimension.
-            expand: The number of dimensions to expand.
-            depth: The number of layers.
-            n_classes: The number of classes.
-            dropout: The dropout rate.
-            layer_norm_eps: The epsilon value for layer normalization.
-            weight_decay: The L2 regularization
-            spectral_norm: Whether to apply spectral normalization to the final linear layer.
-
-        References: 
-            1. Gu, A., & Dao, T. (2023).
->>>>>>> 1e5bf639793de07acee6aaa4da95df9566f29c94
             Mamba: Linear-time sequence modeling with selective state spaces. 
             arXiv preprint arXiv:2312.00752.
         """
@@ -166,7 +122,6 @@ def __init__(self,
 
         self.layer_norm = nn.LayerNorm(d_model, eps=layer_norm_eps)
 
-<<<<<<< HEAD
     def forward(self, x):
         """ Forward pass
         
@@ -175,18 +130,6 @@ def forward(self, x):
         
         Returns: 
             torch.Tensor: the output tensor.
-=======
-    def forward(self, 
-        x: torch.Tensor
-    ) -> torch.Tensor:
-        """ Forward pass of the model.
-
-        Args:
-            x: The input tensor of shape (B, L, D).
-
-        Returns:
-            The output tensor of shape (B, C).
->>>>>>> 1e5bf639793de07acee6aaa4da95df9566f29c94
         """
         x = x.unsqueeze(1).repeat(1, 100, 1)
 
@@ -198,24 +141,4 @@ def forward(self,
 
         x = self.layer_norm(x)  # Final layer normalization
         x = self.fc(x[:, 0, :])
-        return x
-<<<<<<< HEAD
-=======
-    
-    def get_l2_regularization(self):
-        """ Compute the L2 regularization for the model."""
-        l2_reg = 0.0
-        for layer in self.layers:
-            for param in layer.parameters():
-                l2_reg += torch.norm(param, p=2)
-        return layer.weight_decay * l2_reg
->>>>>>> 1e5bf639793de07acee6aaa4da95df9566f29c94
-
-# Usage example:
-# model = Mamba(d_model=16, d_state=16, d_conv=4, expand=2, depth=4)
-# optimizer = torch.optim.AdamW(model.parameters(), lr=0.001, weight_decay=0.01)
-# 
-# # During training:
-# loss = criterion(model(x), y) + model.get_l2_regularization()
-# loss.backward()
-# optimizer.step()
+        return x