diff --git a/llmfoundry/callbacks/sliding_window_size_warmup_callback.py b/llmfoundry/callbacks/sliding_window_size_warmup_callback.py
index a6962c9943..1d2593bbc3 100644
--- a/llmfoundry/callbacks/sliding_window_size_warmup_callback.py
+++ b/llmfoundry/callbacks/sliding_window_size_warmup_callback.py
@@ -3,9 +3,9 @@
 
 from typing import Any
 
-from composer import Callback, Logger, State, TimeUnit
-from git import Optional
+from composer import Logger, State, TimeUnit
 
+from llmfoundry.interfaces import CallbackWithConfig
 from llmfoundry.utils.warnings import experimental_class
 
 __all__ = [
@@ -14,11 +14,12 @@
 
 
 @experimental_class('SlidingWindowSizeWarmerUpper')
-class SlidingWindowSizeWarmerUpper(Callback):
+class SlidingWindowSizeWarmerUpper(CallbackWithConfig):
     """Warms up the sliding window size for the model based on a schedule.
 
     Args:
-        t_warmup (str|None): Warmup duration for sliding window size, defaults to scheduler.t_warmup.
+        train_config (dict[str, Any]): Training configuration.
+        t_warmup (float): Warmup duration (as a fraction of max duration) for sliding window size.
     """
 
     def __init__(
@@ -32,12 +33,18 @@ def __init__(
 
     def before_train_batch(self, state: State, logger: Logger):
         del logger
+        if state.max_duration is None:
+            raise ValueError(
+                'SlidingWindowSizeWarmerUpper callback requires max_duration to be set.',
+            )
         if state.max_duration.unit == TimeUnit.TOKEN:
             current_time_frac = state.timestamp.token.value / state.max_duration.value
         elif state.max_duration.unit == TimeUnit.BATCH:
             current_time_frac = state.timestamp.batch.value / state.max_duration.value
         else:
-            raise ValueError(f'Unsupported time unit {state.max_duration.unit}')
+            raise ValueError(
+                f'Unsupported time unit {state.max_duration.unit=} for SlidingWindowSizeWarmerUpper callback.',
+            )
         if self.orig_sliding_window_size_list is None:
             self.orig_sliding_window_size_list = [None] * len(
                 state.model.model.transformer.blocks,
@@ -50,7 +57,7 @@ def before_train_batch(self, state: State, logger: Logger):
             if self.orig_sliding_window_size_list[idx] is None:
                 self.orig_sliding_window_size_list[
                     idx] = attn_block.sliding_window_size
-            if self.orig_sliding_window_size_list[idx] != -1:
+            if self.orig_sliding_window_size_list[idx] != -1:  # type: ignore
                 attn_block.sliding_window_size = round(
                     max(
                         self.orig_sliding_window_size_list[idx],