Is it possible to continue training on top of an already saved posterior?

[apongos]

For example, say I used infer() to train a posterior via 100,000 simulations, but realize I didn't train it enough. Can I start training on top of that 100,000 simulations and take it to 200,000 simulations without starting all over?

[michaeldeistler]

Please use the flexible interface for that.

from sbi.inference import SNPE, prepare_for_sbi, simulate_for_sbi

simulator, prior = prepare_for_sbi(simulator, prior)
inference = SNPE(prior)

theta, x = simulate_for_sbi(simulator, proposal=prior, num_simulations=1000)
# ==== This is where you want to save the simulations (e.g. with pickle) and load them next time

density_estimator = inference.append_simulations(theta, x).train()
posterior = inference.build_posterior(density_estimator)