diff --git a/src/remove_ai_watermarks/photomaker_restore.py b/src/remove_ai_watermarks/photomaker_restore.py
index 6b6a466..19495d0 100644
--- a/src/remove_ai_watermarks/photomaker_restore.py
+++ b/src/remove_ai_watermarks/photomaker_restore.py
@@ -151,6 +151,14 @@ def _get_pipeline() -> Any:
                 )
             adapter_path = hf_hub_download(repo_id=_PHOTOMAKER_REPO, filename=_PHOTOMAKER_FILE)
             pipe = PhotoMakerStableDiffusionXLPipeline.from_pretrained(_SDXL_MODEL_ID, torch_dtype=dtype)
+            # Move SDXL submodules to the device BEFORE loading the PhotoMaker adapter:
+            # ``load_photomaker_adapter`` reads ``self.device`` / ``self.unet.dtype`` to
+            # place the new ID encoder. If we ``.to(device)`` after, the SDXL submodules
+            # move but the id_encoder stays where it was (custom attribute, not in the
+            # auto-managed module tree), and inference errors with
+            # "Input type (torch.cuda.HalfTensor) and weight type (torch.HalfTensor)
+            # should be the same" (caught empirically 2026-06-04).
+            pipe.to(device)
             # ``pm_version="v1"`` is REQUIRED: the upstream loader defaults to v2 and would
             # build the V2 encoder (PhotoMakerIDEncoder_CLIPInsightfaceExtendtoken), then
             # error on load_state_dict because the v1 weights have a different shape.
@@ -163,8 +171,12 @@ def _get_pipeline() -> Any:
                 trigger_word="img",
                 pm_version="v1",
             )
-            pipe.to(device)
             pipe.fuse_lora()
+            # Belt: also explicitly cast the loaded id_encoder, because some
+            # diffusers/torch combinations leave the encoder buffers untouched even
+            # though ``pipe.to(device)`` ran first.
+            if hasattr(pipe, "id_encoder") and pipe.id_encoder is not None:
+                pipe.id_encoder = pipe.id_encoder.to(device=device, dtype=dtype)
             _pipeline = pipe
     return _pipeline