ComVEX: Computer Vision EXpo
What are the pros?
Every models share similar building objects:
xxxBase: A model's base. For checking common input arguments and storing important variables. Sometimes it can also provide specified weight initialization methods or necessary tensor operations, like patching and flattening images in
xxxBackbone: A model's backbone architecture. It includes every needed components to build the model except the classifier.
xxxBackboneplus a projection head as its classifier. Only accept
xxxConfigas its argument. Similar to Huggingface. Might provide some variants for differenet objective in the future.
xxxConfig: A configuration for all possible coefficients. It also provides model specializations mentioned in the papers.
Consistent Namings for papers and across papers
To make researchers or developers understand implementations as soon as possible, we tightly follow the names of model components from the official papers and be consistent on common namings across papers.
Clear Tensor Operations
We use Einops for almost all tensor operations to unveil the dimensions of tensors, which are usually hidden in the code, and make our implementations explain by themselves.
To expose all possible arguments to users but still remain convenience, we categorize building objects into a hierarchical order with 3 levels listed from bottom to top as below:
Basic : The paper-proposed and essential objects that mostly inherit directly from
nn.Moduleor other Basic objects, like
Wrapper: Intermediate objects or wrappers that organize Basic ones, like
Basic and Model objects are the ones crucial for paper-to-code mappings and model usages, so we require their arguments to be fully explicit to users (list all arguments in
__init__methods). And for the sake of convenience, Wrapper objects can use
**kwargsto pass down necessary arguments. The overall model structures in term of the number of required arguments will look like a hourglass.
- Basic : The paper-proposed and essential objects that mostly inherit directly from
Excluding some common names like
xfor the input tensors,
ff_dropoutfor the dropout rate of feed forward networks, and
act_func_namefor a string of activation function's name supported by PyTorch, all variables, helper functions, and objects should be named meaningfully.
Detailed Model Information
Every models has its own
README.mdthat provides usages, one-by-one argument explanations, and all usable objects and specializations. The official implementations are provided as well if any mentioned in the official paper.
How to install?
pip3 install comvex
How to use?
Please check out the Usage section detailed in models' own
How to contribute?
Please check out the
CONTRIBUTING.md for details.
- Continuously implementing models, please check them out under the
comvexfolder for more details and
examplesfolder for some demos.
- Pull requests are welcome!
- From this issue, inheritance doesn't support in
torchscript. Therefore, most of our implementations aren't scriptable. But
traceseems that doesn't exist this kind of issue and we will use
traceas our default and gradually update our code to make ComVEX a trace-supported library.