The goal of the Mooncake.jl project is to produce an AD package written entirely in Julia that improves on ForwardDiff.jl, ReverseDiff.jl, and Zygote.jl in several ways.
Please refer to the docs for more info.
Important
Mooncake.jl accepts issues and pull requests for reproducible defects only. Feature requests, enhancements, redesign proposals, support requests, and debugging requests without a minimal reproducible example are out of scope and will be closed. Mooncake is intended to differentiate through the vast majority of Julia Base and the standard libraries. Known limitations are documented on the known limitations page. Requests for rules covering functionality outside Julia Base are considered out of scope.
Check that you're running a version of Julia that Mooncake.jl supports.
See the SUPPORT_POLICY.md file for more info.
There are several ways to interact with Mooncake.jl. To interact directly with Mooncake.jl, use Mooncake's native API, which allows reuse of prepared caches for repeated gradient and Hessian evaluations:
import Mooncake as MC
f(x) = (1 - x[1])^2 + 100 * (x[2] - x[1]^2)^2 # Rosenbrock
x = [1.2, 1.2]
# Reverse mode
grad_cache = MC.prepare_gradient_cache(f, x);
val, grad = MC.value_and_gradient!!(grad_cache, f, x)
# Forward mode
fwd_cache = MC.prepare_derivative_cache(f, x);
val_fwd, grad_fwd = MC.value_and_gradient!!(fwd_cache, f, x)
# Hessian
hess_cache = MC.prepare_hessian_cache(f, x);
val, grad, H = MC.value_gradient_and_hessian!!(hess_cache, f, x)
# val : f(x)
# grad : ∇f(x) (length-n vector)
# H : ∇²f(x) (n×n matrix)You should expect that MC.prepare_gradient_cache and MC.prepare_hessian_cache take a little time to run, but that subsequent calls using the prepared caches are fast.
For additional details, see the interface docs. You can also interact with Mooncake.jl via DifferentiationInterface.jl, although this interface may lag behind Mooncake in supporting newly introduced features.