Is there appetite for PRs adding metal kernels of scipy.special functions used in probabilistic programming, such as:
Context: I'm a dev of pymc/pytensor, we're interested in expanding MLX support as a compute backend for bayesian models. These functions are important for computing the logp of various distributions.
Is there appetite for PRs adding metal kernels of scipy.special functions used in probabilistic programming, such as:
Context: I'm a dev of pymc/pytensor, we're interested in expanding MLX support as a compute backend for bayesian models. These functions are important for computing the logp of various distributions.