This branch is up-to-date with the current progress on building REALM, the open domain information retrieval QA system. (We should ensure that this is on a stable branch, ready to use.)
The following sections (will) reflect the three stages of training a REALM system. For now it's just the ICT code.
Loosely, they are pretraining the retriever modules, then jointly training the language model and the retriever, and then finetuning a question answering head on the language model with fixed retriever.
The following sections reflect the three stages of training a REALM system. Loosely, they are pretraining the retriever modules, then jointly training the language model and the retriever, and then finetuning a question answering head on the language model with fixed retriever.
### Inverse Cloze Task (ICT) Pretraining
### Inverse Cloze Task (ICT) Pretraining
1. Have a corpus in loose JSON format with the intention of creating a collection of fixed-size blocks of text as the fundamental units of data. For a corpus like Wikipedia, this will mean multiple sentences per block but also multiple blocks per document.
1. Have a corpus in loose JSON format with the intention of creating a collection of fixed-size blocks of text as the fundamental units of data. For a corpus like Wikipedia, this will mean multiple sentences per block but also multiple blocks per document.
