diff --git a/olmo/checkpoint.py b/olmo/checkpoint.py
index faae84979..1303e151b 100644
--- a/olmo/checkpoint.py
+++ b/olmo/checkpoint.py
@@ -542,7 +542,6 @@ def unshard_checkpoint(
 
         Note this is not marked abstract because child classes are not required to implemented this.
         """
-        del load_path, local_cache, load_optimizer_state, load_trainer_state, device
         raise NotImplementedError
 
     @contextmanager
@@ -1914,13 +1913,13 @@ def unshard_checkpoint(
 
 
 def build_sharded_checkpointer(
-    cfg: TrainConfig, *, name: Optional[ShardedCheckpointerType] = None
+    cfg: TrainConfig, *, name: Optional[ShardedCheckpointerType] = None, use_shared_mem_impl: bool = False
 ) -> Checkpointer:
     name = name or cfg.sharded_checkpointer
     if name == ShardedCheckpointerType.torch_new:
         return TorchNewStyleShardedCheckpointer(cfg)
     elif name == ShardedCheckpointerType.torch_legacy:
-        return TorchLegacyShardedCheckpointer(cfg)
+        return TorchLegacyShardedCheckpointer(cfg, use_shared_mem_impl=use_shared_mem_impl)
     elif name == ShardedCheckpointerType.local:
         return LocalShardedCheckpointer(cfg)
     elif name == ShardedCheckpointerType.olmo_core:
diff --git a/scripts/unshard.py b/scripts/unshard.py
index 1063e2f4f..41cd83c51 100644
--- a/scripts/unshard.py
+++ b/scripts/unshard.py
@@ -1,16 +1,11 @@
 import logging
 import shutil
 from pathlib import Path
-from typing import Union
+from typing import Optional, Union
 
 import torch
 
-from olmo.checkpoint import (
-    Checkpointer,
-    LocalShardedCheckpointer,
-    OlmoCoreCheckpointer,
-    TorchLegacyShardedCheckpointer,
-)
+from olmo.checkpoint import build_sharded_checkpointer
 from olmo.config import ShardedCheckpointerType, TrainConfig
 from olmo.safetensors_util import state_dict_to_safetensors_file
 
@@ -20,7 +15,7 @@
 def main(
     input_dir: Union[str, Path],
     output_dir: Union[str, Path],
-    sharded_checkpoint_type: ShardedCheckpointerType = ShardedCheckpointerType.torch_legacy,
+    sharded_checkpoint_type: Optional[ShardedCheckpointerType] = None,
     model_only: bool = False,
     safe_tensors: bool = False,
     use_shared_mem_impl: bool = False,
@@ -32,15 +27,11 @@ def main(
     output_dir.mkdir(parents=True, exist_ok=True)
 
     config = TrainConfig.load(input_dir / "config.yaml", validate_paths=False)
-    checkpointer: Checkpointer
-    if sharded_checkpoint_type == ShardedCheckpointerType.torch_legacy:
-        checkpointer = TorchLegacyShardedCheckpointer(config, use_shared_mem_impl=use_shared_mem_impl)
-    elif sharded_checkpoint_type == ShardedCheckpointerType.local:
-        checkpointer = LocalShardedCheckpointer(config)
-    elif sharded_checkpoint_type == ShardedCheckpointerType.olmo_core:
-        checkpointer = OlmoCoreCheckpointer(config)
-    else:
-        raise NotImplementedError(sharded_checkpoint_type)
+
+    sharded_checkpoint_type = sharded_checkpoint_type or config.sharded_checkpointer
+    checkpointer = build_sharded_checkpointer(
+        config, name=sharded_checkpoint_type, use_shared_mem_impl=use_shared_mem_impl
+    )
 
     model_state_dict, optim_state_dict, trainer_state_dict = checkpointer.unshard_checkpoint(
         input_dir,
@@ -92,8 +83,8 @@ def main(
     parser.add_argument(
         "--type",
         choices=list(ShardedCheckpointerType),
-        default=ShardedCheckpointerType.torch_legacy,
-        help="""The sharded checkpoint type.""",
+        default=None,
+        help="""The sharded checkpoint type. Defaults to the sharded checkpoint type set in config.""",
     )
     parser.add_argument(
         "--model-only",