Short Summaries
#7
by
shortsnip
- opened
I've found this model also works really well at summarizing shorter documents (100-500 characters). However every now and then something like this happens:
Input: Republicans Are Blaming Biden For A Border Crisis That Republicans Refuse To Fix
Result: Biden is Blaming Biden for a Border Crisis.
Kinda funny but curious if there's anything I can do configuration wise to reduce these kind of occurrences. Thank you!
This model was trained on conversational data.
From your example you should look for models trained to summarize news articles or non-dialogue style texts.
knkarthick
changed discussion status to
closed