better variable naming

aphrodite/modeling/layers/sampler.py

@@ -258,8 +258,8 @@ def _apply_alphabet_soup(
     logits_sort[mask] = -float("inf")
 
     # Apply top-k.
-    for i,k in enumerate(k):
-        logits_sort[i, k:] = -float("inf")
+    for i,topk in enumerate(k):
+        logits_sort[i, topk:] = -float("inf")
 
     # Re-sort the probabilities.
     src = torch.arange(logits_idx.shape[-1],