Crate blazen_train_tune 

AutoML / hyperparameter search on top of blazen-train.

This crate is intentionally separate from blazen-train itself so that the heavy training engine (candle, hf-hub, tokenizers) doesn’t have to be compiled in to use the searchers — the Evaluator trait is the only coupling point, and callers wire their own training loop into it.

Pieces:

• space — SearchSpace + Distribution (Categorical / IntUniform / Uniform / LogUniform / Discrete). The space defines what hyperparams exist and what each one’s prior looks like.
• trial — Trial record + status state machine.
• searcher — Searcher trait + three real implementations: RandomSearch, GridSearch, and TpeSearch (Tree-structured Parzen Estimator, Bergstra et al. 2011, Algorithm 1).
• journal — JSONL append-only TrialJournal for crash recovery.
• runner — Runner glues a SearchSpace + Searcher + Evaluator
  • TrialJournal into a budgeted (max-trials / time-budget) loop, optionally fan-out across N tokio workers.

See examples/lora_sft_search.rs for an end-to-end LoRA SFT search.

ML acronyms (LoRA, TPE, KDE, SFT, …) saturate the docs; backticking each one would hurt readability without adding clarity.

Re-exports

pub use error::TuneError;

pub use journal::TrialJournal;

pub use runner::EvalFuture;

pub use runner::Evaluator;

pub use runner::Runner;

pub use runner::RunnerBudget;

pub use searcher::GridSearch;

pub use searcher::RandomSearch;

pub use searcher::Searcher;

pub use searcher::TpeSearch;

pub use space::Distribution;

pub use space::SearchSpace;

pub use trial::Trial;

pub use trial::TrialId;

pub use trial::TrialStatus;

Modules

error

Error type for blazen-train-tune.

journal

Append-only JSONL trial journal.

runner

The async search loop.

searcher

Search strategies: RandomSearch, GridSearch, TpeSearch.

space

SearchSpace + Distribution types.

trial

Trial record + lifecycle state machine.