TAPAS: Weakly Supervised Table Parsing via Pre-training #4288

[shashankMadan-designEsthetics]

🌟 TAPAS (Weakly Supervised Table Parsing via Pre-Training)
Model description
TAPAS ArXiV Paper Google AI Blog, an approach to question answering over tables without generating logical forms. TAPAS trains from weak supervision, and predicts the denotation by selecting table cells and optionally applying a corresponding aggregation operator to such selection. TAPAS extends BERT’s architecture to encode tables as input, initializes from an effective joint pre-training of text segments and tables crawled from Wikipedia, and is trained end-to-end. We experiment with three different semantic parsing datasets, and find that TAPAS outperforms or rivals semantic parsing models by improving state-of-the-art accuracy on SQA from 55.1 to 67.2 and performing on par with the state-of-the-art on WIKISQL and WIKITQ, but with a simpler model architecture. We additionally find that transfer learning, which is trivial in our setting, from WIKISQL to WIKITQ, yields 48.7 accuracy, 4.2 points above the state-of-the-art.

Open source status
[ X ] the model implementation is available: GitHub repo
[ X ] the model weights are available: in the "Data" section at their GitHub page
[ X ] who are the authors: Jonathan Herzig, Paweł Krzysztof Nowak, Thomas Muller, Francesco Piccinno, Julian Martin Eisenschlos

This integration could be interesting

[stephenroller]

I agree, it would be nice to see this in ParlAI. We don't expect we have resources to prioritize it right now, but we would welcome any contributions toward this goal.