twisterl
| Crates.io | twisterl |
| lib.rs | twisterl |
| version | 0.5.1 |
| created_at | 2025-10-03 09:29:41.790039+00 |
| updated_at | 2026-01-12 09:25:13.166327+00 |
| description | Reinforcement learning primitives and a Python extension for high performance training and inference. |
| homepage | |
| repository | |
| max_upload_size | |
| id | 1866432 |
| size | 91,969 |
(AIforQuantum)
documentation
README
twisterl
twisterl provides the Rust core for the TwisteRL project. It implements reinforcement learning primitives and a Python extension for high performance training and inference.
Features
- Generic
Envtrait for implementing discrete environments - Sample puzzle environment
- Simple neural network layers and policy utilities
- Parallel collectors for PPO and AlphaZero algorithms
- Optional Python bindings via PyO3
This crate is currently a proof of concept and its API may change.
Usage
Add the crate to your Cargo.toml:
[dependencies]
twisterl = { version = "0.1"}
Python interface
When compiled with the python_bindings feature the crate exposes bindings used by the twisterl Python package. See the repository for Python usage examples.
In this case, add the crate to your Cargo.toml with the python_bindings feature:
[dependencies]
twisterl = { version = "0.1", features = ["python_bindings"]}
License
Licensed under the Apache License, Version 2.0..