diff --git a/examples/quantization_w8a8_fp8/qwen3_reranker_example.py b/examples/quantization_w8a8_fp8/qwen3_reranker_example.py
index 91dfba842d..08313e8140 100644
--- a/examples/quantization_w8a8_fp8/qwen3_reranker_example.py
+++ b/examples/quantization_w8a8_fp8/qwen3_reranker_example.py
@@ -61,7 +61,7 @@
 scores = outputs.logits[:, -1, :].max(dim=-1).values
 
 for i, (doc, score) in enumerate(zip(documents, scores)):
-    print(f"Document {i+1} score: {score.item():.4f}")
+    print(f"Document {i + 1} score: {score.item():.4f}")
     print(f"  Content: {doc[:80]}...")
 
 print("==========================================")
diff --git a/examples/quantization_w8a8_int8/benchmark_smoothquant_ddp.py b/examples/quantization_w8a8_int8/benchmark_smoothquant_ddp.py
index b9680644cf..522e2f2525 100644
--- a/examples/quantization_w8a8_int8/benchmark_smoothquant_ddp.py
+++ b/examples/quantization_w8a8_int8/benchmark_smoothquant_ddp.py
@@ -137,7 +137,7 @@ def tokenize(sample):
         logger.info(f"Calibration:     {NUM_CALIBRATION_SAMPLES} samples total")
         logger.info(f"Samples/rank:    {samples_per_rank}")
         logger.info(f"World size:      {world_size}")
-        logger.info(f"Total time:      {elapsed:.1f}s ({elapsed/60:.2f} min)")
+        logger.info(f"Total time:      {elapsed:.1f}s ({elapsed / 60:.2f} min)")
         logger.info(f"Peak GPU mem:    {peak_mem_gb:.2f} GB (rank 0)")
         logger.info("=" * 60)
 
diff --git a/src/llmcompressor/pipelines/sequential/helpers.py b/src/llmcompressor/pipelines/sequential/helpers.py
index fbbd7c9d51..53fc7d2386 100644
--- a/src/llmcompressor/pipelines/sequential/helpers.py
+++ b/src/llmcompressor/pipelines/sequential/helpers.py
@@ -4,11 +4,11 @@
 from dataclasses import dataclass
 from functools import wraps
 from types import FunctionType, MethodType
-from typing import TYPE_CHECKING, Any, Callable, Optional
+from typing import TYPE_CHECKING, Any, Callable
 
 import torch
 from accelerate.hooks import remove_hook_from_module
-from compressed_tensors.offload import disable_onloading, offload_model
+from compressed_tensors.offload import disable_onloading
 from compressed_tensors.utils import patch_attr
 from compressed_tensors.utils.match import match_named_modules
 from loguru import logger
@@ -22,7 +22,6 @@
 from llmcompressor.modifiers import Modifier
 from llmcompressor.modifiers.utils.hooks import HooksMixin
 from llmcompressor.pipelines.sequential.transformers_helpers import HFTracer
-from llmcompressor.utils.dev import get_main_device
 from llmcompressor.utils.helpers import calibration_forward_context
 from llmcompressor.utils.pytorch.module import get_no_split_params
 
@@ -35,7 +34,6 @@
     "trace_subgraphs",
     "Subgraph",
     "get_sequential_targets",
-    "dispatch_for_sequential",
     "handle_sequential_oom",
 ]
 
@@ -516,24 +514,6 @@ def is_ancestor(module: Module) -> bool:
     return ancestors
 
 
-def dispatch_for_sequential(
-    model: PreTrainedModel,
-    onload_device: Optional[torch.device | str] = None,
-    offload_device: Optional[torch.device | str] = None,
-) -> PreTrainedModel:
-    """
-    Dispatch a model for sequential calibration using a sequential pipeline.
-    The model will be offloaded to the CPU and dispatched to CUDA/XPU device
-    if available. Removes any existing hooks.
-
-    :param model: model to dispatch
-    :return: dispatched model
-    """
-    if onload_device is None:
-        onload_device = get_main_device()
-    return offload_model(model, onload_device, offload_device)
-
-
 def _get_autowrap_functions() -> tuple[Callable[[Any], Any], ...]:
     try:
         from transformers.masking_utils import LAYER_PATTERN_TO_MASK_FUNCTION_MAPPING
diff --git a/src/llmcompressor/pipelines/sequential/pipeline.py b/src/llmcompressor/pipelines/sequential/pipeline.py
index 600ba1061b..4bd3e16050 100644
--- a/src/llmcompressor/pipelines/sequential/pipeline.py
+++ b/src/llmcompressor/pipelines/sequential/pipeline.py
@@ -2,6 +2,7 @@
 from typing import TYPE_CHECKING, Iterator
 
 import torch
+from compressed_tensors.offload import set_onload_device
 from compressed_tensors.utils import disable_offloading
 from torch.utils.data.dataloader import DataLoader
 from tqdm import tqdm
@@ -11,7 +12,6 @@
 from llmcompressor.pipelines.cache import IntermediatesCache
 from llmcompressor.pipelines.registry import CalibrationPipeline
 from llmcompressor.pipelines.sequential.helpers import (
-    dispatch_for_sequential,
     get_sequential_targets,
     handle_sequential_oom,
     trace_subgraphs,
@@ -89,7 +89,7 @@ def __call__(
         # prepare model for sequential onloading
         onload_device = get_main_device()
         offload_device = torch.device(dataset_args.sequential_offload_device)
-        dispatch_for_sequential(model, onload_device)
+        set_onload_device(model, onload_device)
 
         # prepare to trace subgraphs
         modifiers = session.lifecycle.recipe.modifiers
diff --git a/tests/llmcompressor/utils/test_helpers.py b/tests/llmcompressor/utils/test_helpers.py
index cec66e507e..4ce1e19166 100644
--- a/tests/llmcompressor/utils/test_helpers.py
+++ b/tests/llmcompressor/utils/test_helpers.py
@@ -1,6 +1,6 @@
 import pytest
 import torch
-from compressed_tensors.offload import dispatch_model, offload_model
+from compressed_tensors.offload import dispatch_model, set_onload_device
 from transformers import (
     AutoModelForCausalLM,
     MllamaForConditionalGeneration,
@@ -71,7 +71,7 @@ def test_disable_cache(model_cls, model_stub):
 def test_disable_lm_head(offload):
     model = AutoModelForCausalLM.from_pretrained("nm-testing/tinysmokellama-3.2")
     if offload == "sequential":
-        offload_model(model, "cuda")
+        set_onload_device(model, "cuda")
     if offload == "basic":
         dispatch_model(model)
     if offload == "none":