Benchmark runner

Installing the package installs a banditry-bench CLI that exercises every agent on synthetic benchmarks:

banditry-bench --agent ofugp-gp    --benchmark branin            --n-iter 30 --seed 42
banditry-bench --agent ts-langevin --benchmark gaussian_valley   --n-iter 30 --seed 42
banditry-bench --agent ofugp-svgp  --benchmark contextual_branin --n-iter 30 --seed 42

Options

Flag	Values	Default
--agent	ofugp-gp, ofugp-svgp, ts-langevin, ts-nuts	ofugp-gp
--benchmark	branin, contextual_branin, gaussian_valley, contextual_gaussian_valley	gaussian_valley
--n-iter	int	30
--seed	int	42
--rand-sample	warmup rounds	4
--noise-std-proxy	float (OFUGP agents only)	1.0
--verbose	flag	off

The contextual_* variants sample one coordinate uniformly each round and pin it via fix_input, exercising the contextual path.

The benchmark functions themselves (branin, gaussian_valley, BENCHMARKS) are importable from banditry.benchmark for reuse in your own scripts and tests.