Datasets:
Dataset split different from decsription
When i download dataset using huggong face load_dataset , the split i am getting is 12460 train, 1500 test, 500 validation but in dataset description it's written 1500 test and 1500 validation, also in paper the split is 12460 train, 500 validation and 500 test, is something wrong with your dataset?
I see it as 13460 as total with the right splits mentioned in the description. 12460 as train , 500 test and 500 validation. I don’t know where exactly are you seeing this error so that I can correct it! Also I have multiple datasets. Could you verify and get back to me!
When i looked at data i found out that you have 3 reference summaries for each test data input dialogue, so it was like each of them as separate data point that's why 500*3 gives 1500 test data points, so it's fine for me now, but thanks for reply