[Bug]: The variable convert_model seems missing in anomalib v1.0.1

[blueclowd]

Describe the bug

The variable convert_model seems missing in anomalib v1.0.1

Code

from anomalib import TaskType
from anomalib.models import Patchcore, Padim, Stfpm, Draem, EfficientAd
from anomalib.engine import Engine
from anomalib.deploy import ExportType
from anomalib.callbacks import ModelCheckpoint
from anomalib.data import Folder
from loguru import logger


# Create the datamodule
datamodule = Folder(
    name="digit7",
    root="/kaggle/input/mykad-7/train_7",
    normal_dir="normal",
    abnormal_dir="abnormal",
    # mask_dir="abnormal_mask",
    task=TaskType.CLASSIFICATION,
    train_batch_size=1
)
datamodule.setup()

model = EfficientAd()
engine = Engine(max_epochs=5, task=TaskType.CLASSIFICATION,callbacks=None)

# Train the model
engine.fit(datamodule=datamodule, model=model)
engine.export(export_type=ExportType.OPENVINO,
              model=model,
              export_root='/kaggle/working/anomalib_weight')

Error message:

File /opt/conda/lib/python3.10/site-packages/anomalib/deploy/export.py:295, in export_to_openvino(export_root, model, transform, ov_args, task)
    293 ov_model_path = model_path.with_suffix(".xml")
    294 ov_args = {} if ov_args is None else ov_args
--> 295 if convert_model is not None and serialize is not None:
    296     model = convert_model(model_path, **ov_args)
    297     serialize(model, ov_model_path)

NameError: name 'convert_model' is not defined

Dataset

N/A

Model

Other (please specify in the field below)

Steps to reproduce the behavior

Code

engine.export(export_type=ExportType.OPENVINO,
              model=model,
              export_root='/kaggle/working/anomalib_weight')

OS information

OS information: Kaggle notebook

• Anomalib version: v1.0.1

Expected behavior

The trained model can be converted to OpenVINO format.

Screenshots

No response

Pip/GitHub

pip

What version/branch did you use?

No response

Configuration YAML

Using default YAML

Logs

Calculate Validation Dataset Quantiles: 100%|██████████| 1/1 [00:00<00:00,  2.13it/s]
Calculate Validation Dataset Quantiles: 100%|██████████| 1/1 [00:00<00:00,  1.92it/s]
Calculate Validation Dataset Quantiles: 100%|██████████| 1/1 [00:00<00:00,  2.13it/s]
Calculate Validation Dataset Quantiles: 100%|██████████| 1/1 [00:00<00:00,  1.91it/s]
Calculate Validation Dataset Quantiles: 100%|██████████| 1/1 [00:00<00:00,  2.11it/s]
INFO: `Trainer.fit` stopped: `max_epochs=5` reached.
/opt/conda/lib/python3.10/site-packages/anomalib/models/image/efficient_ad/torch_model.py:29: TracerWarning: torch.tensor results are registered as constants in the trace. You can safely ignore this warning if you use this function to create tensors out of constant variables that would be the same every time you call this function. In any other case, this might cause the trace to be incorrect.
  mean = torch.tensor([0.485, 0.456, 0.406])[None, :, None, None].to(x.device)
/opt/conda/lib/python3.10/site-packages/anomalib/models/image/efficient_ad/torch_model.py:30: TracerWarning: torch.tensor results are registered as constants in the trace. You can safely ignore this warning if you use this function to create tensors out of constant variables that would be the same every time you call this function. In any other case, this might cause the trace to be incorrect.
  std = torch.tensor([0.229, 0.224, 0.225])[None, :, None, None].to(x.device)
/opt/conda/lib/python3.10/site-packages/anomalib/models/image/efficient_ad/torch_model.py:332: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
  return any(value.sum() != 0 for _, value in p_dic.items())
/opt/conda/lib/python3.10/site-packages/anomalib/models/image/efficient_ad/torch_model.py:212: TracerWarning: Converting a tensor to a Python float might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
  math.ceil(image_size[0] / 4) if self.padding else math.ceil(image_size[0] / 4) - 8,
/opt/conda/lib/python3.10/site-packages/anomalib/models/image/efficient_ad/torch_model.py:213: TracerWarning: Converting a tensor to a Python float might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
  math.ceil(image_size[1] / 4) if self.padding else math.ceil(image_size[1] / 4) - 8,
/opt/conda/lib/python3.10/site-packages/torch/onnx/_internal/jit_utils.py:314: UserWarning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied. (Triggered internally at /usr/local/src/pytorch/torch/csrc/jit/passes/onnx/constant_fold.cpp:179.)
  _C._jit_pass_onnx_node_shape_type_inference(node, params_dict, opset_version)
/opt/conda/lib/python3.10/site-packages/torch/onnx/utils.py:739: UserWarning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied. (Triggered internally at /usr/local/src/pytorch/torch/csrc/jit/passes/onnx/constant_fold.cpp:179.)
  _C._jit_pass_onnx_graph_shape_type_inference(
/opt/conda/lib/python3.10/site-packages/torch/onnx/utils.py:1244: UserWarning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied. (Triggered internally at /usr/local/src/pytorch/torch/csrc/jit/passes/onnx/constant_fold.cpp:179.)
  _C._jit_pass_onnx_graph_shape_type_inference(
---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
Cell In[2], line 33
     29 # logger.info("Checkpoint path: {}".format(engine.trainer.default_root_dir))
     30 
     31 # Train the model
     32 engine.fit(datamodule=datamodule, model=model)
---> 33 engine.export(export_type=ExportType.OPENVINO,
     34               model=model,
     35               export_root='/kaggle/working/anomalib_weight')

File /opt/conda/lib/python3.10/site-packages/anomalib/engine/engine.py:910, in Engine.export(self, model, export_type, export_root, transform, ov_args, ckpt_path)
    903     exported_model_path = export_to_onnx(
    904         model=model,
    905         export_root=export_root,
    906         transform=transform,
    907         task=self.task,
    908     )
    909 elif export_type == ExportType.OPENVINO:
--> 910     exported_model_path = export_to_openvino(
    911         model=model,
    912         export_root=export_root,
    913         transform=transform,
    914         task=self.task,
    915         ov_args=ov_args,
    916     )
    917 else:
    918     logging.error(f"Export type {export_type} is not supported yet.")

File /opt/conda/lib/python3.10/site-packages/anomalib/deploy/export.py:295, in export_to_openvino(export_root, model, transform, ov_args, task)
    293 ov_model_path = model_path.with_suffix(".xml")
    294 ov_args = {} if ov_args is None else ov_args
--> 295 if convert_model is not None and serialize is not None:
    296     model = convert_model(model_path, **ov_args)
    297     serialize(model, ov_model_path)

NameError: name 'convert_model' is not defined

Code of Conduct

• I agree to follow this project's Code of Conduct

[samet-akcay]

Hi @blueclowd, it is probably because you don't have openvino installed on your environment. As you can see below: convert_model is imported from openvino
https://github.com/openvinotoolkit/anomalib/blob/5b01918716de0fb73cc71e1068fd6ba202b26482/src/anomalib/deploy/export.py#L30