Alchemize: Transpile PyMC to Rust for 3-7x speed-up

Дата публикации: 12-05-2026 22:56:37

Excited to release Alchemize: It takes a PyMC model and emits optimized Rust. But the interesting part is what’s not in the codebase: a compiler.
The biggest complexity in PyMC hides in PyTensor: the computational library that takes your high-level Python model description and compiles it to a low-level backend like C, JAX or Numba. Adding a new low-level target like Rust takes a lot of work: every mathematical operation (addition, log, exp) requires its own Rust implementation. PyTensor currently sits at ~150k LOC. Adding a new backend is usually months of work.
But what if instead you let AI write that low-level code directly, bypassing PyTensor all-together? That’s what Alchemize does: About 500 lines of Python scaffolding, a markdown skill file describing how PyMC models map to Rust, and a reference check that compiles the agent’s output, runs it, and compares log-probabilities and gradients against PyMC. The agent writes Rust, validates, self-corrects until the numbers match. Converges in 3-4 iterations. Final Rust runs 3-7x faster than what highly optimized PyTensor produces.
The speedup isn’t the interesting part. The architecture is: Alchemize puts Claude Code (Agent SDK) into a loop that only exits once the compiled Rust produces the exact same outputs. New hardware targets like CUDA or Apple Accelerate don’t need code changes. They only need a new skill file describing what libraries to use.
The compiler isn’t gone. It’s just written in English now.
Code: GitHub - pymc-labs/alchemize: LLM-based, self-correcting transpiler. Supports JAX, PyTorch, Rust, PyMC, Stan. · GitHub
Blog: https://dub.sh/po1I6uu
7 posts - 4 participants
Read full topic

Схожие новости

#	Наименование новости	Тональность	Информативность	Дата публикации
1	🚀 Release v6.3.2	0	18.52	08-09-2026
2	Sampling PyMC models in JupyterLite with a WebAssembly backend for PyTensor	0	6.38	28-09-2026
3	Local sensitivity checks for PyMC posterior predictions (exploratory notebook)	0	5.98	22-03-2026
4	AI Makes "Open-Source, Clean Room" Implementations Of Adobe Photoshop & Premier In Rust	0	9.02	08-10-2026
5	🚀 Release pymc-extras v0.15.1	0	19.63	16-09-2026
6	Hosted MCP server now available	0	6.31	14-09-2026
7	How we make AI coding more cost efficient without sacrificing task quality	0	6.12	02-09-2026
8	Added timestamps for Video 31 (Data Umbrella PyMC Sprint)	0	17.63	09-10-2026
9	Local LLM for Coding: Free AI Coding Agent With Ollama + Claude	0	5.48	11-03-2026
10	Earendil publie Pi 1.0, son harnais d'agent minimal	0	10.7	08-10-2026

Классификация: . Схожих патентов: 0. Схожих новостей: 10. Тональность: 0. Информативность: 7.7. Источник: discourse.pymc.io.