Oneflow-Inc/libai
A distributed training toolbox based on OneFlow for large-scale deep learning model parallelism across data, tensor, and pipeline dimensions.
Not currently ranked — collecting fresh signals.
star history
LiBai is a large-scale open-source model training framework built on OneFlow. It provides multiple parallelism strategies including Data Parallelism, Tensor Parallelism, and Pipeline Parallelism, along with training techniques such as mixed precision training, activation checkpointing, gradient accumulation, and ZeRO optimizer. The toolbox supports both computer vision and NLP tasks with predefined datasets like CIFAR, ImageNet, and BERT.
Frequently asked
- What is Oneflow-Inc/libai?
- A distributed training toolbox based on OneFlow for large-scale deep learning model parallelism across data, tensor, and pipeline dimensions.
- Is libai open source?
- Yes — Oneflow-Inc/libai is open source, released under the Apache-2.0 license.
- What language is libai written in?
- Oneflow-Inc/libai is primarily written in Python.
- How popular is libai?
- Oneflow-Inc/libai has 403 stars on GitHub.
- Where can I find libai?
- Oneflow-Inc/libai is on GitHub at https://github.com/Oneflow-Inc/libai.