Jayesh K. Gupta

While Julia is great, there are still a lot of existing useful differentiable Python code in PyTorch, Jax, etc. Given PyCall.jl is already so great and seamless, one might wonder what it takes to differentiate through those calls to Python functions. PyCallChainRules.jl aims for that ideal. DLPack.jl is leveraged to pass CPU or GPU arrays without any copy between Julia and Python.

Auto differentiation interfaces are rapidly converging with [`functorch`](https://github.com/pytorch/functorch) and [`jax`](https://github.com/google/jax) on the Python side and more explicit interfaces for dealing with gradients in Julia with [`Functors.jl`](https://github.com/FluxML/Functors.jl) and [Optimisers.jl](https://github.com/FluxML/Optimisers.jl) as well as more explicit machine learning layers in [Lux.jl](https://github.com/avik-pal/Lux.jl). While it is relatively easy to implement new functionality in Julia, Python still remains the standard interface layer for most state-of-the-art functionality especially for GPU kernels. Even if not the most performant, there is value in being able to call existing differentiable functions in Python, as a developer slowly implements equivalent functionality in Julia.

Jayesh K. Gupta

Talks:

PyCallChainRules.jl: Reusing differentiable Python code in Julia

Platinum sponsors

Gold sponsors

Silver sponsors

Media partners

Community partners

Fiscal Sponsor