---
license: apache-2.0
task_categories:
- translation
language:
- ko
size_categories:
- 1K<n<10K
---
- Data
  - source
    - MSC data from the paper < Beyond Goldfish Memory: Long-Term Open-Domain Conversation >
    - train/valid/test dataset of session 4 
  - translation ( English -> Koeran )
    - GPT-3.5-turbo is used mostly
    - GPT-4 : 66 data from the start of session_4_train ( after these, changed to gpt-3.5 )