From 0fbdadfe7b39ede46f09b619bd3308db5c8b1476 Mon Sep 17 00:00:00 2001 From: marcoyang Date: Tue, 20 Jun 2023 17:09:52 +0800 Subject: [PATCH] change wording --- .../source/decoding-with-langugage-models/shallow-fusion.rst | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/docs/source/decoding-with-langugage-models/shallow-fusion.rst b/docs/source/decoding-with-langugage-models/shallow-fusion.rst index 3ee8a91ac..e2a3ded67 100644 --- a/docs/source/decoding-with-langugage-models/shallow-fusion.rst +++ b/docs/source/decoding-with-langugage-models/shallow-fusion.rst @@ -3,10 +3,7 @@ Shallow fusion for RNN Transducer ================================= -In real-life scenario, there is often a mismatch between the training corpus and the target corpus space. -Therefore, we often use an external language model (LM) to improve the accuracy of the ASR model -on the target space. Even if the training and testing domain are similar, using external langugage model -can still help the ASR model if the training corpus is not that large. +External language models (LM) are commonly used to improve WERs for E2E ASR models. This tutorial shows you how to perform ``shallow fusion`` with an external LM to improve the word-error-rate of a RNN Transducer model.