ppo2

class Model(object):
    def __init__(self, *, policy, ob_space, ac_space, nbatch_act, nbatch_train,
                nsteps, ent_coef, vf_coef, max_grad_norm, training):
        sess = tf.get_default_session()
        if training:
            act_model = policy(sess, ob_space, ac_space, nbatch_act, 1, reuse=False)
        else:
            act_model = policy(sess, ob_space, ac_space, nbatch_act, 1, reuse=True)
        train_model = policy(sess, ob_space, ac_space, nbatch_train, nsteps, reuse=True)