From 2f4d79470143af51603dc7257f0c5aeba2e91ac9 Mon Sep 17 00:00:00 2001 From: w4ngatang Date: Mon, 10 Jun 2019 10:03:04 -0400 Subject: [PATCH] Update wording of readme --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 43cc108f9..5c254a218 100644 --- a/README.md +++ b/README.md @@ -4,7 +4,7 @@ This branch of `jiant` contains code and scripts to run part of the experiments in the paper [Can You Tell Me How to Get Past Sesame Street? Sentence-Level Pretraining Beyond Language Modeling](https://arxiv.org/abs/1812.10860), which appears at ACL 2019. This paper supersedes "Looking for ELMo's Friends: Sentence-Level Pretraining Beyond Language Modeling", an earlier version of the same paper by the same authors. -Specifically, this branch contains scripts to run experiments in doing additional supervised finetuning of BERT with various NLP tasks (Table 3, BERT with Intermediate Task Training). +Specifically, this branch contains scripts to run experiments that do additional supervised finetuning of BERT with various NLP tasks (Table 3, BERT with Intermediate Task Training). To reproduce the experiments in pretraining sentence encoders from scratch (Table 2) and intermediate training of ELMo (Table 3, ELMo with Intermediate Task Training), use the [`jsalt-experiments`](https://github.com/jsalt18-sentence-repl/jiant/tree/jsalt-experiments) branch. Note that this branch uses a slightly older version of jiant.