Skip to content

[EMNLP2023 main] Just Adjust One Prompt: Enhancing In-Context Dialogue Scoring via Constructing the Optimal Subgraph of Demonstrations and Prompts

Notifications You must be signed in to change notification settings

iamlxb3/EMNLP2023-ADOROR

Repository files navigation

Overview

Part of the code for the ADORER method, we will open source all the code with detailed comments if the paper is accepted.

Code for the core steps

  • prepare_O.py: Code for Algorithm 1, how the data set O is constructed
  • demp_graph.py: Code for Algorithm 2, how we conduct ensemble on the optimal subgraph
  • demo_selector_vanilla_transformer.py: The model structure for f selector with vanilla transformer encoder
  • demo_selector_rel_pos.py: The model structure for f selector with transformer encoder of relation position attention layer

About

[EMNLP2023 main] Just Adjust One Prompt: Enhancing In-Context Dialogue Scoring via Constructing the Optimal Subgraph of Demonstrations and Prompts

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages