Struct DqnTrainer 

pub struct DqnTrainer<E, Enc, Act, B, Buf = CircularBuffer<<E as Environment>::Observation, <E as Environment>::Action>>
where
    E: Environment,
    B: AutodiffBackend,
    Buf: ReplayBuffer<E::Observation, E::Action>,
{ /* private fields */ }

The imperative training driver for DQN.

A thin convenience wrapper around TrainingSession that adds an environment and drives the training loop. Use TrainingSession directly when your loop is owned externally (e.g. Bevy’s ECS).

Usage – iterator style

let mut trainer = DqnTrainer::new(env, agent);

for step in trainer.steps().take(50_000) {
    if step.episode_done {
        println!("Episode {} reward: {}", step.episode, step.episode_reward);
    }
}

Usage – imperative style with run tracking

let mut trainer = DqnTrainer::new(env, agent)
    .with_run(TrainingRun::create("cartpole", "v1")?)
    .with_max_steps(200_000);

trainer.train();
let report = trainer.eval(20);
report.print();

Implementations

impl<E, Enc, Act, B, Buf> DqnTrainer<E, Enc, Act, B, Buf>

where E: Environment, E::Observation: Clone + Send + Sync + 'static, E::Action: Clone + Send + Sync + 'static, Enc: ObservationEncoder<E::Observation, B> + ObservationEncoder<E::Observation, B::InnerBackend>, Act: DiscreteActionMapper<E::Action>, B: AutodiffBackend, Buf: ReplayBuffer<E::Observation, E::Action>,

pub fn new(env: E, agent: DqnAgent<E, Enc, Act, B, Buf>) -> Self

pub fn with_run(self, run: TrainingRun) -> Self

Attach a TrainingRun for checkpoint saving and JSONL logging.

pub fn with_max_steps(self, n: usize) -> Self

Maximum training steps. train() stops when this is reached.

pub fn with_checkpoint_freq(self, freq: usize) -> Self

Checkpoint frequency in steps. Default: 10_000.

pub fn with_keep_checkpoints(self, keep: usize) -> Self

Number of numbered checkpoints to keep on disk. Default: 3.

pub fn with_stats(self, stats: StatsTracker) -> Self

Replace the default stats tracker.

pub fn steps(&mut self) -> TrainIter<'_, E, Enc, Act, B, Buf>

Returns an iterator that yields StepMetrics after each environment step.

pub fn train(&mut self)

Run until max_steps is reached (or forever if not set).

pub fn eval(&mut self, n_episodes: usize) -> EvalReport

Run n_episodes of greedy evaluation and return an EvalReport.

pub fn session(&self) -> &TrainingSession<E, DqnAgent<E, Enc, Act, B, Buf>>

Read-only access to the underlying session.

pub fn into_agent(self) -> DqnAgent<E, Enc, Act, B, Buf>

Consume the trainer and return the inner agent.

pub fn env(&self) -> &E

Access the environment directly.