open-mmlab/mmengine
OpenMMLab Foundational Library for Training Deep Learning Models
This foundational library helps deep learning engineers efficiently train and validate their PyTorch models. It takes your model code and dataset definitions, streamlining the training process to produce optimized models ready for deployment. The target users are deep learning engineers and researchers who build and experiment with various deep learning architectures and need robust training infrastructure.
1,456 stars. Used by 12 other packages. Available on PyPI.
Use this if you are a deep learning engineer or researcher building models with PyTorch and need a flexible, powerful framework to manage model training, integrate large-scale training strategies, and monitor experiments.
Not ideal if you are a data scientist or analyst primarily using pre-built machine learning libraries for tasks like tabular data analysis without custom deep learning model development.
Stars
1,456
Forks
446
Language
Python
License
Apache-2.0
Category
Last pushed
Dec 23, 2025
Commits (30d)
0
Dependencies
9
Reverse dependents
12
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/open-mmlab/mmengine"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related frameworks
Xilinx/brevitas
Brevitas: neural network quantization in PyTorch
google/qkeras
QKeras: a quantization deep learning library for Tensorflow Keras
fastmachinelearning/qonnx
QONNX: Arbitrary-Precision Quantized Neural Networks in ONNX
tensorflow/model-optimization
A toolkit to optimize ML models for deployment for Keras and TensorFlow, including quantization...
lucidrains/vector-quantize-pytorch
Vector (and Scalar) Quantization, in Pytorch