user@devops:~$ cat README.md
Model Export & Inference
# Description
Project focused on ML model export and deployment for production. Covers the full cycle: train a model in Keras, export in multiple formats (SavedModel for TensorFlow Serving, portable HDF5, TF-Lite for mobile devices, ONNX for interoperability), optimize with quantization (float16, int8), and create a Flask REST API that loads the model and serves real-time predictions. Includes a Python test client and endpoint documentation.
# Key features
$ Export in 4 formats: SavedModel, HDF5, TF-Lite, ONNX
$ Optimization with float16 and int8 quantization for production
$ Flask REST API for real-time prediction serving
$ Test client with requests for endpoint validation
$ Image preprocessing integrated in inference pipeline
$ API documentation with usage examples
# Gallery
# Technologies used
