Module model 

Context models — predict next bit probability from context.

Phase 2: Order-0 model (256-context partial byte predictor). Phase 3: Order-1/2/3 context models, match model, CM engine.

Re-exports

pub use cm_model::AssociativeContextModel;

pub use cm_model::ChecksumContextModel;

pub use cm_model::ContextModel;

pub use dmc_model::DmcModel;

pub use engine::CMConfig;

pub use engine::CMEngine;

pub use gru_model::GruModel;

pub use match_model::MatchModel;

pub use neural_model::NeuralModel;

pub use order0::Order0Model;

pub use ppm_model::PpmConfig;

pub use ppm_model::PpmModel;

pub use run_model::RunModel;

pub use sparse_model::SparseModel;

pub use word_model::WordModel;

Modules

cm_model

ContextModel – flexible context model using ContextMap + StateMap.

dmc_model

DMC (Dynamic Markov Compression) — bit-level automaton predictor.

engine

CMEngine – orchestrates all context models + mixer + APM.

gru_model

GRU (Gated Recurrent Unit) byte-level predictor with truncated BPTT.

indirect_model

IndirectModel — second-order context prediction.

json_model

JsonModel – JSON structure-aware context model.

match_model

MatchModel – ring buffer + hash table for longest match prediction.

neural_model

NeuralModel – bit-level cross-context model.

order0

Order-0 Context Model — predicts next bit from partial byte context.

ppm_model

PPM (Prediction by Partial Matching) byte-level predictor — PPMd variant.

run_model

RunModel – run-length context model.

sparse_model

SparseModel – skip-context model for periodic patterns.

word_model

WordModel – word boundary context model.