TensorFlow Serving docker image - that runs on Macbook M1 Pro (under emulation).
As of Jan 2022, TF serving official docker image won't run on M1 Apple machines even with the --platform linux/amd64 option, likely because of emulation issues.
Using the tensorflow-model-server-universal package fixes this problem and allows emulation. As a tradeoff, it also disables some optimizations. Not recommended in a production environment.
Only tested on my MacBook M1 Pro, zero guarantees included.
# Download the TensorFlow Serving Docker image and repo
docker pull cedricbl/tf-serving-universal-amd64
git clone https://github.com/tensorflow/serving
# Location of demo models
TESTDATA="$(pwd)/serving/tensorflow_serving/servables/tensorflow/testdata" # bash
# set TESTDATA (pwd)"/serving/tensorflow_serving/servables/tensorflow/testdata" # fish
# Start TensorFlow Serving container and open the REST API port
docker run -t --rm -p 8501:8501 \
-v "$TESTDATA/saved_model_half_plus_two_cpu:/models/half_plus_two" \
-e MODEL_NAME=half_plus_two \
--platform linux/amd64 \
cedricbl/tf-serving-universal-amd64 &
# Query the model using the predict API
curl -d '{"instances": [1.0, 2.0, 5.0]}' \
-X POST http://localhost:8501/v1/models/half_plus_two:predict
# Returns => { "predictions": [2.5, 3.0, 4.5] }docker build --platform linux/amd64 -t cedricbl/tf-serving-universal-amd64 .