flexs.baselines.explorers.adalead

Defines the Adalead explorer class.

class flexs.baselines.explorers.adalead.Adalead(model, rounds, sequences_batch_size, model_queries_per_batch, starting_sequence, alphabet, mu=1, recomb_rate=0, threshold=0.05, rho=0, eval_batch_size=20, log_file=None)

Bases: flexs.explorer.Explorer

Adalead explorer.

Algorithm works as follows:

Initialize set of top sequences whose fitnesses are at least

(1 - threshold) of the maximum fitness so far

While we can still make model queries in this batch

Recombine top sequences and append to parents Rollout from parents and append to mutants

propose_sequences(measured_sequences)

Propose top sequences_batch_size sequences for evaluation.

Return type

Tuple[ndarray, ndarray]