Update for nncf_task

README.md

@@ -126,11 +126,11 @@ python tools/inference.py \
     --image_path datasets/MVTec/bottle/test/broken_large/000.png
 ```
 
-If you want to run OpenVINO model, ensure that `compression` `apply` is set to `True` in the respective model `config.yaml`.
+If you want to run OpenVINO model, ensure that `openvino` `apply` is set to `True` in the respective model `config.yaml`.
 
 ```yaml
 optimization:
-  compression:
+  openvino:
     apply: true
 ```
 

anomalib/models/__init__.py

@@ -23,6 +23,10 @@
 
 from anomalib.models.components import AnomalyModule
 
+# TODO(AlexanderDokuchaev): Workaround of wrapping by NNCF.
+#                           Can't not wrap `spatial_softmax2d` if use import_module.
+from anomalib.models.padim.model import PadimLightning  # noqa: F401
+
 
 def get_model(config: Union[DictConfig, ListConfig]) -> AnomalyModule:
     """Load model from the configuration file.

anomalib/models/ganomaly/config.yaml

@@ -46,24 +46,8 @@ project:
   save_to_csv: false
 
 optimization:
-  compression:
+  openvino:
     apply: false
-  nncf:
-    apply: false
-    input_info:
-      sample_size: null
-    compression:
-      algorithm: quantization
-      initializer:
-        range:
-          num_init_samples: 256
-    update_config:
-      init_weights: snapshot.ckpt
-      hyperparameter_search:
-        parameters:
-          lr:
-            min: 1e-4
-            max: 1e-2
 
 # PL Trainer Args. Don't add extra parameter here.
 trainer:

anomalib/models/padim/config.yaml

@@ -39,20 +39,8 @@ project:
   save_to_csv: false
 
 optimization:
-  compression:
+  openvino:
     apply: false
-  nncf:
-    apply: false
-    input_info:
-      sample_size: [1, 3, 256, 256]
-    compression:
-      algorithm: quantization
-      initializer:
-        range:
-          num_init_samples: 256
-      ignored_scopes: []
-    update_config:
-      init_weights: snapshot.ckpt
 
 # PL Trainer Args. Don't add extra parameter here.
 trainer:

anomalib/models/padim/model.py

@@ -219,7 +219,8 @@ def smooth_anomaly_map(self, anomaly_map: Tensor) -> Tensor:
         """
 
         kernel_size = 2 * int(4.0 * self.sigma + 0.5) + 1
-        anomaly_map = gaussian_blur2d(anomaly_map, (kernel_size, kernel_size), sigma=(self.sigma, self.sigma))
+        sigma = torch.as_tensor(self.sigma).to(anomaly_map.device)
+        anomaly_map = gaussian_blur2d(anomaly_map, (kernel_size, kernel_size), sigma=(sigma, sigma))
 
         return anomaly_map
 

anomalib/models/stfpm/config.yaml

@@ -46,24 +46,8 @@ project:
   save_to_csv: false
 
 optimization:
-  compression:
+  openvino:
     apply: false
-  nncf:
-    apply: false
-    input_info:
-      sample_size: null
-    compression:
-      algorithm: quantization
-      initializer:
-        range:
-          num_init_samples: 256
-    update_config:
-      init_weights: snapshot.ckpt
-      hyperparameter_search:
-        parameters:
-          lr:
-            min: 1e-4
-            max: 1e-2
 
 # PL Trainer Args. Don't add extra parameter here.
 trainer:

anomalib/utils/callbacks/__init__.py

@@ -18,19 +18,20 @@
 from importlib import import_module
 from typing import List, Union
 
-from omegaconf import DictConfig, ListConfig
+import yaml
+from omegaconf import DictConfig, ListConfig, OmegaConf
 from pytorch_lightning.callbacks import Callback, ModelCheckpoint
 
 from .cdf_normalization import CdfNormalizationCallback
-from .compress import CompressModelCallback
 from .min_max_normalization import MinMaxNormalizationCallback
 from .model_loader import LoadModelCallback
+from .openvino import OpenVINOCallback
 from .save_to_csv import SaveToCSVCallback
 from .timer import TimerCallback
 from .visualizer_callback import VisualizerCallback
 
 __all__ = [
-    "CompressModelCallback",
+    "OpenVINOCallback",
     "LoadModelCallback",
     "TimerCallback",
     "VisualizerCallback",
@@ -69,10 +70,9 @@ def get_callbacks(config: Union[ListConfig, DictConfig]) -> List[Callback]:
     if "normalization_method" in config.model.keys() and not config.model.normalization_method == "none":
         if config.model.normalization_method == "cdf":
             if config.model.name in ["padim", "stfpm"]:
-                if not config.optimization.nncf.apply:
-                    callbacks.append(CdfNormalizationCallback())
-                else:
+                if "nncf" in config.optimization and config.optimization.nncf.apply:
                     raise NotImplementedError("CDF Score Normalization is currently not compatible with NNCF.")
+                callbacks.append(CdfNormalizationCallback())
             else:
                 raise NotImplementedError("Score Normalization is currently supported for PADIM and STFPM only.")
         elif config.model.normalization_method == "min_max":
@@ -84,24 +84,24 @@ def get_callbacks(config: Union[ListConfig, DictConfig]) -> List[Callback]:
         callbacks.append(VisualizerCallback(inputs_are_normalized=not config.model.normalization_method == "none"))
 
     if "optimization" in config.keys():
-        if config.optimization.nncf.apply:
+        if "nncf" in config.optimization and config.optimization.nncf.apply:
             # NNCF wraps torch's jit which conflicts with kornia's jit calls.
             # Hence, nncf is imported only when required
-            nncf_module = import_module("anomalib.utils.callbacks.nncf_callback")
+            nncf_module = import_module("anomalib.utils.callbacks.nncf.callback")
             nncf_callback = getattr(nncf_module, "NNCFCallback")
+            nncf_config = yaml.safe_load(OmegaConf.to_yaml(config.optimization.nncf))
             callbacks.append(
                 nncf_callback(
-                    config=config,
-                    dirpath=os.path.join(config.project.path, "compressed"),
-                    filename="compressed_model",
+                    config=nncf_config,
+                    export_dir=os.path.join(config.project.path, "compressed"),
                 )
             )
-        if config.optimization.compression.apply:
+        if "openvino" in config.optimization and config.optimization.openvino.apply:
             callbacks.append(
-                CompressModelCallback(
+                OpenVINOCallback(
                     input_size=config.model.input_size,
-                    dirpath=os.path.join(config.project.path, "compressed"),
-                    filename="compressed_model",
+                    dirpath=os.path.join(config.project.path, "openvino"),
+                    filename="openvino_model",
                 )
             )
 

anomalib/utils/callbacks/nncf/__init__.py

@@ -0,0 +1,15 @@
+"""Integration NNCF."""
+
+# Copyright (C) 2021 Intel Corporation
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions
+# and limitations under the License.

anomalib/utils/callbacks/nncf/callback.py

@@ -0,0 +1,98 @@
+"""Callbacks for NNCF optimization."""
+
+# Copyright (C) 2022 Intel Corporation
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions
+# and limitations under the License.
+
+import os
+from typing import Any, Dict, Optional
+
+import pytorch_lightning as pl
+from nncf import NNCFConfig
+from nncf.api.compression import CompressionAlgorithmController
+from nncf.torch import register_default_init_args
+from pytorch_lightning import Callback
+
+from anomalib.utils.callbacks.nncf.utils import InitLoader, wrap_nncf_model
+
+
+class NNCFCallback(Callback):
+    """Callback for NNCF compression.
+
+    Assumes that the pl module contains a 'model' attribute, which is
+    the PyTorch module that must be compressed.
+
+    Args:
+        config (Dict): NNCF Configuration
+        export_dir (Str): Path where the export `onnx` and the OpenVINO `xml` and `bin` IR are saved.
+                          If None model will not be exported.
+    """
+
+    def __init__(self, nncf_config: Dict, export_dir: str = None):
+        self.export_dir = export_dir
+        self.nncf_config = NNCFConfig(nncf_config)
+        self.nncf_ctrl: Optional[CompressionAlgorithmController] = None
+
+    # pylint: disable=unused-argument
+    def setup(self, trainer: pl.Trainer, pl_module: pl.LightningModule, stage: Optional[str] = None) -> None:
+        """Call when fit or test begins.
+
+        Takes the pytorch model and wraps it using the compression controller
+        so that it is ready for nncf fine-tuning.
+        """
+        if self.nncf_ctrl is not None:
+            return
+
+        init_loader = InitLoader(trainer.datamodule.train_dataloader())  # type: ignore
+        nncf_config = register_default_init_args(self.nncf_config, init_loader)
+
+        self.nncf_ctrl, pl_module.model = wrap_nncf_model(
+            model=pl_module.model, config=nncf_config, dataloader=trainer.datamodule.train_dataloader()  # type: ignore
+        )
+
+    def on_train_batch_start(
+        self,
+        trainer: pl.Trainer,
+        _pl_module: pl.LightningModule,
+        _batch: Any,
+        _batch_idx: int,
+        _unused: Optional[int] = 0,
+    ) -> None:
+        """Call when the train batch begins.
+
+        Prepare compression method to continue training the model in the next step.
+        """
+        if self.nncf_ctrl:
+            self.nncf_ctrl.scheduler.step()
+
+    def on_train_epoch_start(self, _trainer: pl.Trainer, _pl_module: pl.LightningModule) -> None:
+        """Call when the train epoch starts.
+
+        Prepare compression method to continue training the model in the next epoch.
+        """
+        if self.nncf_ctrl:
+            self.nncf_ctrl.scheduler.epoch_step()
+
+    def on_train_end(self, _trainer: pl.Trainer, _pl_module: pl.LightningModule) -> None:
+        """Call when the train ends.
+
+        Exports onnx model and if compression controller is not None, uses the onnx model to generate the OpenVINO IR.
+        """
+        if self.export_dir is None or self.nncf_ctrl is None:
+            return
+
+        os.makedirs(self.export_dir, exist_ok=True)
+        onnx_path = os.path.join(self.export_dir, "model_nncf.onnx")
+        self.nncf_ctrl.export_model(onnx_path)
+        optimize_command = "mo --input_model " + onnx_path + " --output_dir " + self.export_dir
+        os.system(optimize_command)