diff --git a/llama_stack/providers/inline/post_training/__init__.py b/llama_stack/providers/inline/post_training/__init__.py
new file mode 100644
index 000000000..756f351d8
--- /dev/null
+++ b/llama_stack/providers/inline/post_training/__init__.py
@@ -0,0 +1,5 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
diff --git a/llama_stack/providers/inline/post_training/torchtune/common/__init__.py b/llama_stack/providers/inline/post_training/torchtune/common/__init__.py
new file mode 100644
index 000000000..756f351d8
--- /dev/null
+++ b/llama_stack/providers/inline/post_training/torchtune/common/__init__.py
@@ -0,0 +1,5 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
diff --git a/llama_stack/providers/inline/post_training/torchtune/common/utils.py b/llama_stack/providers/inline/post_training/torchtune/common/utils.py
index 2b7a4ec93..b4cd43770 100644
--- a/llama_stack/providers/inline/post_training/torchtune/common/utils.py
+++ b/llama_stack/providers/inline/post_training/torchtune/common/utils.py
@@ -16,8 +16,6 @@ from typing import Any, Callable, Dict, List
 import torch
 from llama_models.datatypes import Model
 from llama_models.sku_list import resolve_model
-from llama_stack.apis.common.type_system import ParamType, StringType
-from llama_stack.apis.datasets import Datasets
 
 from pydantic import BaseModel
 
@@ -26,6 +24,9 @@ from torchtune.models.llama3._tokenizer import Llama3Tokenizer
 from torchtune.models.llama3_1 import lora_llama3_1_8b
 from torchtune.models.llama3_2 import lora_llama3_2_3b
 
+from llama_stack.apis.common.type_system import ParamType, StringType
+from llama_stack.apis.datasets import Datasets
+
 
 class ColumnName(Enum):
     instruction = "instruction"
diff --git a/llama_stack/providers/inline/post_training/torchtune/datasets/__init__.py b/llama_stack/providers/inline/post_training/torchtune/datasets/__init__.py
new file mode 100644
index 000000000..756f351d8
--- /dev/null
+++ b/llama_stack/providers/inline/post_training/torchtune/datasets/__init__.py
@@ -0,0 +1,5 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
diff --git a/llama_stack/providers/inline/post_training/torchtune/recipes/__init__.py b/llama_stack/providers/inline/post_training/torchtune/recipes/__init__.py
new file mode 100644
index 000000000..756f351d8
--- /dev/null
+++ b/llama_stack/providers/inline/post_training/torchtune/recipes/__init__.py
@@ -0,0 +1,5 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
diff --git a/llama_stack/providers/inline/post_training/torchtune/recipes/lora_finetuning_single_device.py b/llama_stack/providers/inline/post_training/torchtune/recipes/lora_finetuning_single_device.py
index a2ef1c5dd..6c795d310 100644
--- a/llama_stack/providers/inline/post_training/torchtune/recipes/lora_finetuning_single_device.py
+++ b/llama_stack/providers/inline/post_training/torchtune/recipes/lora_finetuning_single_device.py
@@ -14,6 +14,24 @@ from typing import Any, Dict, List, Optional, Tuple
 
 import torch
 from llama_models.sku_list import resolve_model
+from torch import nn
+from torch.optim import Optimizer
+from torch.utils.data import DataLoader, DistributedSampler
+from torchtune import modules, training, utils as torchtune_utils
+from torchtune.data import AlpacaToMessages, padded_collate_sft
+
+from torchtune.modules.loss import CEWithChunkedOutputLoss
+from torchtune.modules.peft import (
+    get_adapter_params,
+    get_adapter_state_dict,
+    get_lora_module_names,
+    get_merged_lora_ckpt,
+    set_trainable_params,
+    validate_missing_and_unexpected_for_lora,
+)
+from torchtune.training.lr_schedulers import get_cosine_schedule_with_warmup
+from torchtune.training.metric_logging import DiskLogger
+from tqdm import tqdm
 
 from llama_stack.apis.common.training_types import PostTrainingMetric
 from llama_stack.apis.datasetio import DatasetIO
@@ -38,24 +56,6 @@ from llama_stack.providers.inline.post_training.torchtune.config import (
     TorchtunePostTrainingConfig,
 )
 from llama_stack.providers.inline.post_training.torchtune.datasets.sft import SFTDataset
-from torch import nn
-from torch.optim import Optimizer
-from torch.utils.data import DataLoader, DistributedSampler
-from torchtune import modules, training, utils as torchtune_utils
-from torchtune.data import AlpacaToMessages, padded_collate_sft
-
-from torchtune.modules.loss import CEWithChunkedOutputLoss
-from torchtune.modules.peft import (
-    get_adapter_params,
-    get_adapter_state_dict,
-    get_lora_module_names,
-    get_merged_lora_ckpt,
-    set_trainable_params,
-    validate_missing_and_unexpected_for_lora,
-)
-from torchtune.training.lr_schedulers import get_cosine_schedule_with_warmup
-from torchtune.training.metric_logging import DiskLogger
-from tqdm import tqdm
 
 log = logging.getLogger(__name__)