Project has been completed.
Project description
How to setup airflow
Set airflow directory
export AIRFLOW_HOME="/home/avnish/census_consumer_project/census_consumer_complaint/airflow"
To install airflow
pip install apache-airflow
To configure databse
airflow db init
To create login user for airflow
airflow users create -e avnish@ineuron.ai -f Avnish -l Yadav -p admin -r Admin -u admin
To start scheduler
airflow scheduler
To launch airflow server
airflow webserver -p <port_number>
pip install pandas-tfrecords
pip install \
--upgrade --ignore-installed \
python-snappy==0.5.1 \
--global-option=build_ext \
--global-option="-I/usr/local/include" \
--global-option="-L/usr/local/lib"
pip install twine python setup.py sdist bdist_wheel twine upload --repository-url https://test.pypi.org/legacy/ dist/* twine upload dist/*
To deploy your model
pip install tensorflow-serving-api
to inspect model
saved_model_cli show --dir <dir_path>
Above command will return tag set
saved_model_cli show --dir <dir_path> --tag_set <tag_name>
Above command will show available model signatures
Next: with tag_set and signature_def info, we can inspect model input and output
saved_model_cli show --dir <dir_path> --tag_set <tag_name> --signature_def <SignatureDef Key>
To inspect all signature without tag_set and signature_def saved_model_cli show --dir <dir_path> --all
Testing the model
Test model prediction using saved_model_cli with sample input data
--input_examples: input data formatted as a tf.Example data structure
other param
--outdir: by default output will be written in terminal
--overwrite: to write into a file
tf_debug: run in debug mode
To expose your model as an API using docker image tensorflow/serving
docker pull tensorflow/serving
Single model configuration
sudo docker run -p 8500:8500 \
-p 8501:8501\
--volumn <model_dir>:<target_dir>\
-e MODEL_NAME=<model_name>\
-e model_base_path=<target_dir>\
-t tensorflow/serving:latest
sudo docker run -p 8500:8500 -p 8501:8501 \
-v /home/avnish/census_consumer_project/census_consumer_complaint/census_consumer_complaint_data/saved_models:/avnish/my_model \
-e MODEL_NAME=my_model \
-e MODEL_BASE_PATH=/avnish \
-t tensorflow/serving:latest
To inspect docker container directory
docker exec -it <conatiner_name> bash
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for census-consumer-complaint-0.1.0.tar.gz
Algorithm | Hash digest | |
---|---|---|
SHA256 | f6c7f369709911f7df7924dfaae1a3f5204ec1b322b413abb0408223b67bbcd7 |
|
MD5 | be1a830b03852196696b05c8e823b7fe |
|
BLAKE2b-256 | 50ea04f4087ee590fb8a932b0a7cb624c5d2146deec235f2bc0c6acd535b53a5 |
Hashes for census_consumer_complaint-0.1.0-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 11d8d5e80d0f15b4463efece6af1ca78a12527afe03d8ee6b8be7658bc206a80 |
|
MD5 | 8b127e56d21dd595c8f1eaee4675097d |
|
BLAKE2b-256 | 3b1edc72e6154ea750ca459b79751b70b41e077fa087de936acb3a10df4d1c3a |