make format

netam/common.py

@@ -110,11 +110,10 @@ def codon_mask_tensor_of(nt_parent, *other_nt_seqs, aa_length=None):
 
 
 def aa_strs_from_idx_tensor(idx_tensor):
-    """
-    Convert a tensor of amino acid indices back to a list of amino acid strings.
-    
+    """Convert a tensor of amino acid indices back to a list of amino acid strings.
+
     Args:
-        idx_tensor (Tensor): A 2D tensor of shape (batch_size, seq_len) containing 
+        idx_tensor (Tensor): A 2D tensor of shape (batch_size, seq_len) containing
                              indices into AA_STR_SORTED_AMBIG.
 
     Returns:
@@ -126,7 +125,7 @@ def aa_strs_from_idx_tensor(idx_tensor):
     for row in idx_tensor:
         aa_str = "".join(AA_STR_SORTED_AMBIG[idx] for idx in row.tolist())
         aa_str_list.append(aa_str.rstrip("X"))
-    
+
     return aa_str_list
 
 

netam/models.py

@@ -730,23 +730,23 @@ def predict(self, representation: Tensor):
 class TransformerBinarySelectionModelPIE(TransformerBinarySelectionModelWiggleAct):
     """Here the beta parameter is fixed at 0.3."""
 
-    def __init__(self, 
-                 esm_model_name: str, 
-                 layer_count: int,
-                 dropout_prob: float = 0.5,
-                 output_dim: int = 1,
-                 ):
+    def __init__(
+        self,
+        esm_model_name: str,
+        layer_count: int,
+        dropout_prob: float = 0.5,
+        output_dim: int = 1,
+    ):
         self.pie = ESMEmbedder(model_name=esm_model_name)
         super().__init__(
             nhead=self.pie.num_heads,
             d_model_per_head=self.pie.d_model_per_head,
             # TODO this is hard coded as per Antoine.
-            dim_feedforward=self.pie.d_model*4,
+            dim_feedforward=self.pie.d_model * 4,
             layer_count=layer_count,
             dropout_prob=dropout_prob,
             output_dim=output_dim,
         )
-
 
     def to(self, device):
         super().to(device)
@@ -767,9 +767,9 @@ def represent(self, amino_acid_indices: Tensor, mask: Tensor) -> Tensor:
             where E is the dimensionality of the embedding space.
         """
         # Multiply by sqrt(d_model) to match the transformer paper.
-        embedded_amino_acids = self.pie.embed_batch(
-            amino_acid_indices
-        ) * math.sqrt(self.d_model)
+        embedded_amino_acids = self.pie.embed_batch(amino_acid_indices) * math.sqrt(
+            self.d_model
+        )
         # Have to do the permutation because the positional encoding expects the
         # sequence length to be the first dimension.
         embedded_amino_acids = self.pos_encoder(

netam/protein_embedders.py

@@ -6,8 +6,7 @@
 
 
 def pad_embeddings(embeddings, desired_length):
-    """
-    Pads a batch of embeddings to a specified sequence length with zeros.
+    """Pads a batch of embeddings to a specified sequence length with zeros.
 
     Args:
         embeddings (torch.Tensor): Input tensor of shape (batch_size, seq_len, embedding_dim).
@@ -36,8 +35,7 @@ def pad_embeddings(embeddings, desired_length):
 
 class ESMEmbedder:
     def __init__(self, model_name: str, device: str = "cpu"):
-        """
-        Initializes the ESMEmbedder object.
+        """Initializes the ESMEmbedder object.
 
         Args:
             model_name (str): Name of the pretrained ESM model (e.g., "esm2_t6_8M_UR50D").
@@ -65,8 +63,7 @@ def num_layers(self):
         return self.model.num_layers
 
     def embed_sequence_list(self, sequences: list[str]) -> torch.Tensor:
-        """
-        Embeds a batch of sequences.
+        """Embeds a batch of sequences.
 
         Args:
             sequences (list[str]): List of amino acid sequences.
@@ -84,8 +81,7 @@ def embed_sequence_list(self, sequences: list[str]) -> torch.Tensor:
         return embeddings
 
     def embed_batch(self, amino_acid_indices: torch.Tensor) -> torch.Tensor:
-        """
-        Embeds a batch of netam amino acid indices.
+        """Embeds a batch of netam amino acid indices.
 
         For now, we detokenize the amino acid indices and then use embed_sequence_list.