Salesforce
/

qa_consolidation

Text Classification

question_answering

answer_consolidation

Inference Endpoints

Model card Files Files and versions Community

philippelaban commited on Oct 17, 2022

Commit

f18d231

·

1 Parent(s): 58b4d5c

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -12,7 +12,7 @@ widget:
 ---
 # QA Consolidation Model
-Model card for the QA Consolidation (step 3) of the Discord Questions framework (EMNLP 2022 - Findings).
 The model is a RoBERTa-large model, finetuned on the [MOCHA dataset](https://arxiv.org/abs/2010.03636), and a 5-pt version of the [Answer Equivalence](https://arxiv.org/abs/2202.07654v1) dataset. For a (question, answer1, answer2)-tuple, the model outputs a [1-5] answer similarity score, where 5 is most similar.

 ---
 # QA Consolidation Model
+Model card for the QA Consolidation (step 3) of the Discord Questions framework (EMNLP 2022 - Findings). The model assesses the similarity between two answers (a1, a2) to a question Q. The score obtained is on a scale from 1 (most dissimilar) to 5 (most similar). See example below for formatting.
 The model is a RoBERTa-large model, finetuned on the [MOCHA dataset](https://arxiv.org/abs/2010.03636), and a 5-pt version of the [Answer Equivalence](https://arxiv.org/abs/2202.07654v1) dataset. For a (question, answer1, answer2)-tuple, the model outputs a [1-5] answer similarity score, where 5 is most similar.