Machine Learning Training Utilities for TensorFlow 2.* and PyTorch with Python 3
To use MLTU in your own project, you can install it from PyPI:
pip install mltu
When running tutorials, it's necessary to install mltu for a specific tutorial, for example:
pip install mltu==0.1.3
Each tutorial has its own requirements.txt file for a specific mltu version. As this project progress, the newest versions may have breaking changes, so it's recommended to use the same version as in the tutorial.
Tutorials and Examples can be found on PyLessons.com
- Text Recognition With TensorFlow and CTC network, code in
Tutorials\01_image_to_word
folder; - TensorFlow OCR model for reading Captchas, code in
Tutorials\02_captcha_to_text
folder; - Handwriting words recognition with TensorFlow, code in
Tutorials\03_handwriting_recognition
folder; - Handwritten sentence recognition with TensorFlow, code in
Tutorials\04_sentence_recognition
folder; - Introduction to speech recognition with TensorFlow, code in
Tutorials\05_speech_recognition
folder; - Introduction to PyTorch in a practical way, code in
Tutorials\06_pytorch_introduction
folder; - Using custom wrapper to simplify PyTorch models training pipeline, code in
Tutorials\07_pytorch_wrapper
folder; - Handwriting words recognition with PyTorch, code in
Tutorials\08_handwriting_recognition_torch
folder; - Transformer training with TensorFlow for Translation task, code in
Tutorials\09_translation_transformer
folder;