Getting import error

I am trying to install Intel Extension for PyTorch in a Docker container. I get error:

  File "/usr/src/app/.../Predictor.py", line 4, in <module>
    import intel_extension_for_pytorch as ipex
  File "/usr/local/lib/python3.8/site-packages/intel_extension_for_pytorch/__init__.py", line 13, in <module>
    from . import quantization
  File "/usr/local/lib/python3.8/site-packages/intel_extension_for_pytorch/quantization/__init__.py", line 1, in <module>
    from .quantization_utils import calibrate, convert
  File "/usr/local/lib/python3.8/site-packages/intel_extension_for_pytorch/quantization/quantization_utils.py", line 6, in <module>
    import intel_extension_for_pytorch._C as core
ImportError: /usr/local/lib/python3.8/site-packages/intel_extension_for_pytorch/lib/libintel-ext-pt-cpu.so: undefined symbol: _ZNK3c104Type14isSubtypeOfExtERKS0_PSo
make: *** [Makefile:6: ssp_train] Error 1

My Dockerfile:

FROM python:3.8

ENV POETRY_VIRTUALENVS_CREATE=false \
    POETRY_VERSION=1.1.12 \
    PYTHONDONTWRITEBYTECODE=1 \
    PYTHONUNBUFFERED=1

WORKDIR /usr/src/app
RUN apt-get update
RUN apt-get --yes install libsnappy-dev libgomp1

COPY ./poetry.lock ./poetry.lock
COPY ./pyproject.toml ./pyproject.toml

RUN pip install poetry
RUN poetry install --no-interaction

COPY . .
RUN cp .env.example .env

CMD ["python3", "-m", "train_model"]

I am using poetry for package management and pip for package installation. How can I fix this error? I have only added intel-extension-for-pytorch = "1.11.*" to my dependencies, other packages are e.g. numpy, scipy, pytorch (regular pip versions).

[j-adamczyk]

The model is MLP in PyTorch, 3 hidden layers, learning rate 0.01, dropout 0.3. Inputs are text embeddings from DistilBERT, length 768 dense vectors. Layer shapes: 768 (input), 64, 32, 1 (regression target).

CPU: Intel® Core™ i7-10850H CPU @ 2.70GHz × 12

[j-adamczyk]

MSELoss