diff --git a/Dockerfile-noGPU b/Dockerfile-noGPU
deleted file mode 100644
index 916869b..0000000
--- a/Dockerfile-noGPU
+++ /dev/null
@@ -1,38 +0,0 @@
-FROM docker.io/python:3.10
-
-# Create dirs for:
-# - Injecting config.yml: /root/.DANE
-# - Mount point for input & output files: /mnt/dane-fs
-# - Storing the source code: /src
-RUN mkdir /root/.DANE /mnt/dane-fs /src /model
-
-
-WORKDIR /src
-
-# copy the pyproject file and install all the dependencies first
-RUN pip install --upgrade pip
-RUN pip install poetry
-COPY ./pyproject.toml /src
-RUN --mount=type=cache,target=/home/.cache/pypoetry/cache \
-    --mount=type=cache,target=/home/.cache/pypoetry/artifacts \
-    poetry config virtualenvs.create false && \
-    poetry install --only main --no-interaction --no-ansi
-
-# copy the rest into the source dir
-COPY ./ /src
-
-# Write provenance info about software versions to file
-RUN echo "dane-visual-feature-extraction-worker;https://github.com/beeldengeluid/dane-visual-feature-extraction-worker/commit/$(git rev-parse HEAD)" >> /software_provenance.txt
-
-ENTRYPOINT ["./docker-entrypoint.sh"]
-
-# NOTE:  RUN pip install --no-cache-dir torch should drop the image size
-
-# OF: && pip cache purge (na poetry install command)
-
-# syntax=docker/dockerfile:1.2
-
-# COPY poetry.lock /
-# RUN --mount=type=cache,target=/home/.cache/pypoetry/cache \
-#     --mount=type=cache,target=/home/.cache/pypoetry/artifacts \
-#     poetry install
\ No newline at end of file
diff --git a/README.md b/README.md
index c240bae..c5f5c94 100644
--- a/README.md
+++ b/README.md
@@ -7,13 +7,11 @@ Apply VisXP models trained for feature extraction from keyframes and audio spect
 
 From the root of the repo, run 
 ```
-DOCKER_BUILDKIT=1 docker build -t dane-visual-feature-extraction-worker . 
+docker build -t dane-visual-feature-extraction-worker . 
 ```
-Buildkit is optional, it may speed up building (see https://docs.docker.com/build/buildkit/)
 NB: building the image has occasionally led to mysterious connection errors, which we haven't been able to track down and solve (yet). Discarding the poetry.lock has been a way to circumvent these. 
 
-The default Dockerfile is based on an image that allows GPU processing. A Dockerfile based on a regular Python base image is provided in `Dockerfile-noGPU`, but should not be essential. 
-Whether the processing actually uses GPU depends on the availability of GPU in the container. 
+The Dockerfile support both CPU and GPU processing. Whether the processing actually uses GPU depends on the availability of GPU in the container. 
 
 ### Config