Index

_

__init__() (gradhpo.algorithms.baselines.FOOptimizer method) (gradhpo.algorithms.baselines.OneStepOptimizer method) (gradhpo.algorithms.greedy.GreedyOptimizer method) (gradhpo.algorithms.online.OnlineHypergradientOptimizer method) (gradhpo.algorithms.t1t2.T1T2Optimizer method) (gradhpo.core.base.BilevelOptimizer method) (gradhpo.core.state.BilevelState method)

B

BilevelOptimizer (class in gradhpo.core.base)

BilevelState (class in gradhpo.core.state)

C

compute_hypergradient() (gradhpo.algorithms.baselines.FOOptimizer method) (gradhpo.algorithms.baselines.OneStepOptimizer method) (gradhpo.algorithms.greedy.GreedyOptimizer method) (gradhpo.algorithms.online.OnlineHypergradientOptimizer method) (gradhpo.algorithms.t1t2.T1T2Optimizer method) (gradhpo.core.base.BilevelOptimizer method)

create() (gradhpo.core.state.BilevelState class method)

D

DataBatch (class in gradhpo.core.types)

E

estimate_theta() (gradhpo.algorithms.online.OnlineHypergradientOptimizer method)

F

FOOptimizer (class in gradhpo.algorithms.baselines)

G

get_metric() (gradhpo.core.state.BilevelState method) gradhpo.algorithms.baselines module gradhpo.algorithms.greedy module gradhpo.algorithms.online module gradhpo.algorithms.t1t2 module

gradhpo.core.base module gradhpo.core.state module gradhpo.core.types module gradhpo.utils.gradients module GreedyOptimizer (class in gradhpo.algorithms.greedy)

H

hyperparams (gradhpo.core.state.BilevelState attribute)

I

init() (gradhpo.algorithms.baselines.FOOptimizer method) (gradhpo.algorithms.baselines.OneStepOptimizer method) (gradhpo.algorithms.greedy.GreedyOptimizer method) (gradhpo.algorithms.online.OnlineHypergradientOptimizer method) (gradhpo.algorithms.t1t2.T1T2Optimizer method) (gradhpo.core.base.BilevelOptimizer method)

inner_opt_state (gradhpo.core.state.BilevelState attribute) inputs (gradhpo.core.types.DataBatch attribute)

L

LossFunctions (class in gradhpo.core.types)

M

metadata (gradhpo.core.state.BilevelState attribute) module gradhpo.algorithms.baselines gradhpo.algorithms.greedy gradhpo.algorithms.online gradhpo.algorithms.t1t2 gradhpo.core.base gradhpo.core.state gradhpo.core.types gradhpo.utils.gradients

O

OneStepOptimizer (class in gradhpo.algorithms.baselines)

OnlineHypergradientOptimizer (class in gradhpo.algorithms.online) outer_opt_state (gradhpo.core.state.BilevelState attribute)

P

params (gradhpo.core.state.BilevelState attribute)

R

run() (gradhpo.algorithms.baselines.FOOptimizer method) (gradhpo.algorithms.baselines.OneStepOptimizer method) (gradhpo.algorithms.greedy.GreedyOptimizer method) (gradhpo.algorithms.online.OnlineHypergradientOptimizer method) (gradhpo.algorithms.t1t2.T1T2Optimizer method)

S

step (gradhpo.core.state.BilevelState attribute) step() (gradhpo.algorithms.baselines.FOOptimizer method) (gradhpo.algorithms.baselines.OneStepOptimizer method) (gradhpo.algorithms.greedy.GreedyOptimizer method) (gradhpo.algorithms.online.OnlineHypergradientOptimizer method) (gradhpo.algorithms.t1t2.T1T2Optimizer method) (gradhpo.core.base.BilevelOptimizer method)

T

T1T2Optimizer (class in gradhpo.algorithms.t1t2) targets (gradhpo.core.types.DataBatch attribute) train_loss (gradhpo.core.types.LossFunctions attribute) tree_dot() (in module gradhpo.utils.gradients)

tree_l2_norm() (in module gradhpo.utils.gradients) tree_lerp() (in module gradhpo.utils.gradients) tree_normalize() (in module gradhpo.utils.gradients) tree_zeros_like() (in module gradhpo.utils.gradients)

U

update() (gradhpo.core.state.BilevelState method)

update_w_star() (in module gradhpo.utils.gradients)

V

val_loss (gradhpo.core.types.LossFunctions attribute)

vjp_wrt_both() (in module gradhpo.utils.gradients) vjp_wrt_lambda() (in module gradhpo.utils.gradients)