본문 바로가기

Reference

Blenderbot 2.0: Long-term memory module

ParlAI/modules.py at 0d23cf1cd7b1dac1bdc72f0a852cad5525b95d0c · facebookresearch/ParlAI · GitHub

GitHub - facebookresearch/ParlAI: A framework for training and evaluating AI models on a variety of openly available dialogue da

A framework for training and evaluating AI models on a variety of openly available dialogue datasets. - GitHub - facebookresearch/ParlAI: A framework for training and evaluating AI models on a vari...

github.com

Long term memory module writes document embeddings to a memory + retrieval then scores all documents in the memory and returns the final results

Long term memory module은 memory에 document embedding을 저장하고 retrieval을 통해 memory의 document를 scoring 한 후 final result를 출력한다.

[Base logic]

  1. Retrieve
    1. Retrieve from FAISS or search
    2. Memory search
  2. Expand the input

[Base scheme]

'Reference' 카테고리의 다른 글

[Dataset] Multi Session Dataset list-up  (0) 2024.06.20