Index

A

aepo module aepo() (in module aepo.cli) aepo.cli module aepo.mbr module aepo.mbr.mbr_engine module aepo.mbr.policy module

aepo.mbr.policy.diverse_mbr module aepo.mbr.policy.mbr module aepo.mbr.reward_engine module aepo.mbr.reward_model module aepo.mbr.utility_func module aepo.preprocess module

C

compute_dmbr() (in module aepo.mbr.policy.diverse_mbr)

compute_mbr() (in module aepo.mbr.policy.mbr) compute_score_matrix() (in module aepo.mbr.policy.mbr)

D

ds2csv() (in module aepo.preprocess)

E

Eurus (class in aepo.mbr.reward_model)

F

forward() (aepo.mbr.reward_model.GPTRewardModel method)

G

gbfs() (in module aepo.mbr.policy.diverse_mbr) generate_objective() (in module aepo.mbr.policy.diverse_mbr) get_device() (aepo.mbr.reward_model.GPTRewardModel method) get_pairwise_reward() (aepo.mbr.reward_model.PairLM method) (aepo.mbr.reward_model.RewardModel method) (aepo.mbr.reward_model.StanfordNLP method) get_pairwise_rewards() (aepo.mbr.reward_model.RewardModel method) get_reward() (aepo.mbr.reward_model.Eurus method) (aepo.mbr.reward_model.OASST method) (aepo.mbr.reward_model.PairLM method) (aepo.mbr.reward_model.RewardModel method) (aepo.mbr.reward_model.StanfordNLP method) (aepo.mbr.reward_model.Starling method)

get_reward_() (aepo.mbr.reward_model.Starling method) get_rewards() (aepo.mbr.reward_model.PairLM method) (aepo.mbr.reward_model.RewardModel method) (aepo.mbr.reward_model.Starling method) get_winratio() (aepo.mbr.reward_model.PairLM method) GPTRewardModel (class in aepo.mbr.reward_model)

L

load_reward_model() (in module aepo.mbr.reward_model)

load_similarity() (in module aepo.mbr.utility_func) local_search() (in module aepo.mbr.policy.diverse_mbr)

M

module aepo aepo.cli aepo.mbr aepo.mbr.mbr_engine aepo.mbr.policy aepo.mbr.policy.diverse_mbr aepo.mbr.policy.mbr aepo.mbr.reward_engine aepo.mbr.reward_model aepo.mbr.utility_func aepo.preprocess

O

OASST (class in aepo.mbr.reward_model)

P

PairLM (class in aepo.mbr.reward_model)

R

read_dataset() (in module aepo.preprocess) RewardModel (class in aepo.mbr.reward_model)

run_mbr() (in module aepo.mbr.mbr_engine) run_reward() (in module aepo.mbr.reward_engine)

S

StanfordNLP (class in aepo.mbr.reward_model)

Starling (class in aepo.mbr.reward_model)

T

tokenize_pair() (aepo.mbr.reward_model.PairLM method)