add step_limit to dynamicqueryattention (#159)

This just adds a way to have a given attention mechanism stop before it
has completed all steps.
This makes sense because even though an attention mechanism is trained
for n steps we may not always need to go the whole n steps.
I'm using this right now to train it for classification on the average
of all steps. I think it should make it to shimmer since we'll probably
use this functionality quite a lot.

shimmer/modules/selection.py

@@ -225,9 +225,23 @@ def __init__(
             {domain: nn.Linear(domain_dim, head_size) for domain in domain_names}
         )
         self.n_steps = n_steps
+        self.step_limit = n_steps  # Default step limit is n_steps
         # Start with a random gw state
         self.register_buffer("initial_gw_state", torch.rand(domain_dim))
 
+    def set_step_limit(self, step_limit: int):
+        """
+        Sets the step limit for the dynamic attention update loop.
+
+        Args:
+            step_limit (`int`): Maximum number of steps to run the loop.
+        """
+        if step_limit > self.n_steps:
+            raise ValueError(
+                f"Step limit cannot exceed the maximum n_steps ({self.n_steps})."
+            )
+        self.step_limit = step_limit
+
     def fuse_weighted_encodings(
         self, encodings: LatentsDomainGroupT, attention_dict: dict[str, torch.Tensor]
     ) -> torch.Tensor:
@@ -289,7 +303,7 @@ def forward(
 
         if self.n_steps > 0:
             # Update the query based on the static attention scores
-            for _ in range(self.n_steps):
+            for _ in range(min(self.step_limit, self.n_steps)):
                 # Apply the attention scores to the encodings
                 summed_tensor = self.fuse_weighted_encodings(
                     encodings_pre_fusion, attention_dict