Hello all,
To help you ace this challenge, we are sharing a list of resources for Task 1 and Task 2.
LLM leaderboard Open LLM Leaderboard - a Hugging Face Space by HuggingFaceH4
- participants can choose and change LLM from “Baseline LLM-Phi2-Prompt” which is currently using Phi-2 (microsoft/phi-2 · Hugging Face)
- it’s better to choose models with about 7B or less parameters in terms of resource restrictions
- participants can change LLM, modify prompts, and/or fine-tune LLMs
Dialogue dataset & models
- Topcial-Chat
- GitHub - alexa/Topical-Chat: A dataset containing human-human knowledge-grounded open-domain conversations.
- https://arxiv.org/pdf/2308.11995.pdf
- OpenAssistant
- GitHub - LAION-AI/Open-Assistant: OpenAssistant is a chat-based assistant that understands tasks, can interact with third-party systems, and retrieve information dynamically to do so.
- https://arxiv.org/pdf/2304.07327.pdf
- LMSYS-CHAT-1M
- GitHub - lm-sys/FastChat: An open platform for training, serving, and evaluating large language models. Release repo for Vicuna and Chatbot Arena.
- https://arxiv.org/pdf/2309.11998.pdf
Harry Potter Dialogue (HPD)
Commonsense Knowledge Graph (CSKG):
- PeaCoK (Persona-Grounded CSKG):
- Comet-Atomic 2020 (General Social CSKG, Knowledge Model)
- Persona-Grounded Dialogue Modeling:
- PersonaChat:
ArXiv: [1801.07243] Personalizing Dialogue Agents: I have a dog, do you have pets too?
- PersonaChat:
- PersonaChat Leaderboard (LMEDR, P2Bot, TransferTransfo):
Persona-Chat Benchmark (Dialogue Generation) | Papers With Code - PPG (Partner Persona Induction):
Commonsense Knowledge Construction
- ComFact (Contextual Commonsense Linking Benchmark):
- ArXiv: [2210.12678] ComFact: A Benchmark for Linking Contextual Commonsense Knowledge
GitHub (original, relation-aware): GitHub - Silin159/ComFact
GitHub (relation-agnostic): GitHub - Silin159/ComFact-Relation-Agnostic
- ArXiv: [2210.12678] ComFact: A Benchmark for Linking Contextual Commonsense Knowledge
- Kogito (Commonsense Knowledge Inference Toolkit):
Make your first submission for Prompt Engineer Track in Task 1 with the help of this easy to follow video: https://www.youtube.com/watch?v=5I2DoCiG0kc
Do these resources give you new ideas for your solution? How will you use these resources? Share your ideas or questions down below! 