# big_vision: Google Research's Codebase for Large-Scale Vision Models This repository profile is provided by osrepos.com, an open source repository discovery platform. Source: osrepos.com Repository profile: https://osrepos.com/repo/google-research-big_vision Generated for open source discovery and AI-assisted research. big_vision is Google Research's official codebase for training large-scale vision models using Jax/Flax. It has been instrumental in developing prominent architectures like Vision Transformer, SigLIP, and MLP-Mixer. This repository offers a robust starting point for researchers to conduct scalable vision experiments on GPUs and Cloud TPUs, scaling seamlessly from single cores to distributed setups. GitHub: https://github.com/google-research/big_vision OSRepos URL: https://osrepos.com/repo/google-research-big_vision ## Summary big_vision is Google Research's official codebase for training large-scale vision models using Jax/Flax. It has been instrumental in developing prominent architectures like Vision Transformer, SigLIP, and MLP-Mixer. This repository offers a robust starting point for researchers to conduct scalable vision experiments on GPUs and Cloud TPUs, scaling seamlessly from single cores to distributed setups. ## Topics - Python - Jupyter Notebook - Machine Learning - Computer Vision - Deep Learning - Vision Transformer - Google Research - AI ## Repository Information Last analyzed by OSRepos: Wed Dec 31 2025 00:00:59 GMT+0000 (Western European Standard Time) Detail views: 13 GitHub clicks: 12 ## Safety Notice OSRepos shares public repositories for knowledge and discovery only. Review source code, dependencies, licenses, and security implications before running or installing anything. ## Content ## Introduction big_vision is the official codebase from Google Research, designed for training large-scale vision models. Built upon the powerful Jax and Flax libraries, it leverages tf.data and TensorFlow Datasets for creating scalable and reproducible input pipelines. This repository serves two primary purposes: to publish the code for various research projects developed within Google, and to provide a robust starting point for researchers to conduct large-scale vision experiments. It supports seamless scaling from a single GPU core to distributed setups with up to 2048 Cloud TPU cores. Notable architectures developed using big_vision include Vision Transformer (ViT), SigLIP, MLP-Mixer, and LiT. While the codebase is dynamic, its core functionality is maintained, and it aims to support Google's internal research, with external contributions generally requiring pre-approval. ## Installation To get started with big_vision on a GPU machine, follow these steps. It is highly recommended to use a virtual environment for dependency management. 1. **Clone the Repository:** bash git clone https://github.com/google-research/big_vision cd big_vision/ 2. **Install Python Dependencies:** bash pip3 install --upgrade pip pip3 install -r big_vision/requirements.txt 3. **Install JAX with CUDA Support:** bash pip3 install --upgrade "jax[cuda]" -f https://storage.googleapis.com/jax-releases/jax_cuda_releases.html *Note: You may need to adjust the JAX package based on your specific CUDA and cuDNN versions. Refer to the official JAX documentation for more details.* 4. **Prepare TFDS Data:** big_vision uses `tensorflow_datasets` for unified and reproducible access to standard datasets. It's recommended to prepare datasets separately before running experiments. For example, to download and preprocess `cifar100`, `oxford_iiit_pet`, and `imagenet_v2`: bash cd big_vision/ python3 -m big_vision.tools.download_tfds_datasets cifar100 oxford_iiit_pet imagenet_v2 Some datasets, like `imagenet2012`, require manual download of raw data files into `$TFDS_DATA_DIR/downloads/manual/` before running the `download_tfds_datasets` command. ## Examples Once installed and data is prepared, you can run training jobs. Here are examples for GPU machines: * **Train ViT-S/16 on ImageNet:** bash python3 -m big_vision.train --config big_vision/configs/vit_s16_i1k.py --workdir workdirs/`date '+%m-%d_%H%M'` * **Train MLP-Mixer-B/16 (with GPU-specific batch size):** bash python3 -m big_vision.train --config big_vision/configs/mlp_mixer_i1k.py:gpu8 --workdir workdirs/`date '+%m-%d_%H%M'` The repository's README also provides detailed instructions and commands for running experiments on Google Cloud TPU VMs, including multi-host setups and FSDP training. ## Why Use big_vision? big_vision offers several compelling advantages for researchers and developers working with large-scale vision models: * **Scalability:** Designed for high-performance training, it scales effortlessly from single GPU machines to massive distributed setups on Cloud TPUs, supporting up to 2048 TPU cores. * **Research Foundation:** It is the foundational codebase for numerous cutting-edge research projects from Google, including Vision Transformer, MLP-Mixer, LiT, and SigLIP, providing battle-tested implementations. * **Robustness:** Training jobs are robust to interruptions, capable of seamlessly resuming from the last saved checkpoint, ensuring reliability for long-running experiments. * **Powerful Configuration System:** Features a flexible configuration system that allows for easy customization and extension of training parameters and modules. * **Reproducibility:** Utilizes `tf.data` and `TensorFlow Datasets` to ensure scalable and reproducible input pipelines, crucial for scientific research. ## Links * **GitHub Repository:** [https://github.com/google-research/big_vision](https://github.com/google-research/big_vision) * **ViT Baseline Paper:** [Better plain ViT baselines for ImageNet-1k](https://arxiv.org/abs/2205.01580) * **JAX Documentation:** [https://github.com/jax-ml/jax#pip-installation-gpu-cuda](https://github.com/jax-ml/jax#pip-installation-gpu-cuda) * **TensorFlow Datasets Catalog:** [https://www.tensorflow.org/datasets/catalog/overview#all_datasets](https://www.tensorflow.org/datasets/catalog/overview#all_datasets)