LeapLabTHU/EfficientTrain
1.5−3.0× lossless training or pre-training speedup. An off-the-shelf, easy-to-implement algorithm for the efficient training of foundation visual backbones.
This project offers an algorithm called EfficientTrain++ that significantly speeds up the training process for visual AI models. It takes your existing visual datasets and model architectures as input and produces the same high-accuracy models in less time, making it ideal for AI researchers and machine learning engineers developing computer vision applications.
226 stars. No commits in the last 6 months.
Use this if you need to train or pre-train visual backbone models faster without losing any accuracy, especially for large datasets like ImageNet-1K/22K.
Not ideal if you are not working with visual foundation models or are not concerned with optimizing training speed.
Stars
226
Forks
9
Language
Python
License
MIT
Category
Last pushed
Aug 23, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/LeapLabTHU/EfficientTrain"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
open-mmlab/mmengine
OpenMMLab Foundational Library for Training Deep Learning Models
Xilinx/brevitas
Brevitas: neural network quantization in PyTorch
fastmachinelearning/qonnx
QONNX: Arbitrary-Precision Quantized Neural Networks in ONNX
google/qkeras
QKeras: a quantization deep learning library for Tensorflow Keras
tensorflow/model-optimization
A toolkit to optimize ML models for deployment for Keras and TensorFlow, including quantization...