Crate rlox_nn 

rlox-nn: Backend-agnostic neural network traits for reinforcement learning.

Defines the ActorCritic, QFunction, StochasticPolicy, DeterministicPolicy, and EntropyTuner traits that any NN backend (Candle, Burn, or custom) can implement.

Key types:

• TensorData: Flat Vec<f32> with shape metadata, safe for FFI.
• ActionOutput, EvalOutput: Structured policy outputs.
• PPOStepConfig, DQNStepConfig: Algorithm hyperparameters.

Re-exports

pub use error::NNError;

pub use tensor_data::TensorData;

pub use traits::*;

Modules

distributions

Pure-Rust distribution utilities for CPU-based sampling and log-prob. These are backend-independent helpers that can be used by any backend or for testing without a NN framework.

error

tensor_data

traits