rlkit/torch/sac/bear.py [14:43]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    def __init__(
        self,
        env,
        policy,
        qf1,
        qf2,
        target_qf1,
        target_qf2,
        vae,

        discount=0.99,
        reward_scale=1.0,

        policy_lr=1e-3,
        qf_lr=1e-3,
        optimizer_class=optim.Adam,

        soft_target_tau=1e-2,
        target_update_period=1,
        plotter=None,
        render_eval_paths=False,

        # BEAR specific params
        mode='auto',
        kernel_choice='laplacian',
        policy_update_style=0,
        mmd_sigma=10.0,
        target_mmd_thresh=0.05,
        num_samples_mmd_match=4,
        use_target_nets=True,
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -



rlkit/torch/sac/uwac_dropout.py [19:48]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    def __init__(
        self,
        env,
        policy,
        qf1,
        qf2,
        target_qf1,
        target_qf2,
        vae,

        discount=0.99,
        reward_scale=1.0,

        policy_lr=1e-3,
        qf_lr=1e-3,
        optimizer_class=optim.Adam,

        soft_target_tau=1e-2,
        target_update_period=1,
        plotter=None,
        render_eval_paths=False,

        # BEAR specific params
        mode='auto',
        kernel_choice='laplacian',
        policy_update_style=0,
        mmd_sigma=10.0,
        target_mmd_thresh=0.05,
        num_samples_mmd_match=4,
        use_target_nets=True,
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -