| ### Dialogue Summarizations | |
| General formats: | |
| ```js | |
| { | |
| "dataset_name--train/val/test--dialog_id": { | |
| "original dialog id": str, | |
| "dialog index": int, | |
| "original dialog info": { | |
| "summary": str, | |
| }, | |
| "log": [ | |
| { | |
| "turn id": int, | |
| "user utterance": str, | |
| "system response": str, | |
| "dialog history": str, | |
| "original user side information": dict, | |
| "original system side information": dict, | |
| }, | |
| ... | |
| ], | |
| "prompt": [ | |
| "This is a conversation between two speakers. Given the dialogue context, please generate a summarization about the dialogue.", | |
| ... | |
| ] | |
| }, | |
| ``` | |
| Notice that we cannot form a turn with utterances from only two sides for those datasets consisting of multi-party dialogs (e.g. AMI, ICSI). Therefore, we dump the dialog context and summary in the "original dialog info" and leave the "log" blank. | |