Datasets:
Tasks:
Text Classification
Sub-tasks:
natural-language-inference
Languages:
English
Multilinguality:
monolingual
Size Categories:
1K<n<10K
Language Creators:
crowdsourced
Annotations Creators:
expert-generated
Source Datasets:
original
License:
why id-3 is true?
#1
by
noneUsername
- opened
"Sumit is hungry. Yann is thirsty. Yann travelled to the kitchen. Jason is bored." no anything about "garden".
What task ? I find the agent motivation original task sloppy, if you're mentioning this one.
What task ? I find the agent motivation original task sloppy, if you're mentioning this one.
yes, babi_nli\agents-motivations\train.tsv, the 5 line:
Sumit is hungry. Yann is thirsty. Yann travelled to the kitchen. Jason is bored. Jason will go to the garden. 1
You can look at the original babi paper for proper description. I only generated NEGATIVE examples to have entailment/not entailement (see colab)
That said, I excluded this particular subtask (agent-motivation) from tasksource because I found it ill-defined.