+
Skip to content
/ jimm Public

JAX Image Modeling of Models contains Computer Vision/Vision Language Model implementations in native flax nnx with proper sharding annotations (allowing you to do easy fully sharded data parallel training with native flax nnx/jax)

License

Notifications You must be signed in to change notification settings

Locamage/jimm

Repository files navigation

Jax Image Modeling of Models (jimm)

Docs are at: https://locamage.dev/jimm

  • This aims to be the jax counterpart to timm, with the exception that for image-text models (CLIP, SigLIP, etc), we support the text model entirely.
  • Made with flax nnx, supports weight loading from pytorch_model.bin and safetensors (as well as both methods from huggingface).

Models Supported:

  • Vision Transformers
    • Both with a classification linear layer, or not
    • Using a CLS Token for pooling, or using Multihead Attention Pooling
    • Can load any standard variant of Vision Transformers of any size/resolution(e.g. "google/vit-base-patch16-224" or "google/vit-large-patch16-384")
  • CLIP
    • Can load from any checkpoints of the clip model on github (such as "openai/clip-vit-base-patch32" or "geolocal/StreetCLIP")
  • SigLIP
    • Can load any non-naflex version of the SigLIP model, from both siglipv1 and siglipv2 (eg "google/siglip-base-patch16-256" or "google/siglip2-large-patch16-512" from huggingface or locally)

Installation

Using pixi.sh:

pixi add timm@https://github.com/Locamage/jimm.git --pypi

Using uv

uv add --dev git+https://github.com/Locamage/jimm.git or if you prefer to not add as a direct dependency: uv pip install git+https://github.com/Locamage/jimm.git

Using pip/conda

pip install git+https://github.com/Locamage.git

About

JAX Image Modeling of Models contains Computer Vision/Vision Language Model implementations in native flax nnx with proper sharding annotations (allowing you to do easy fully sharded data parallel training with native flax nnx/jax)

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Contributors 5

Languages

点击 这是indexloc提供的php浏览器服务,不要输入任何密码和下载